A Performance Analysis of View Maintenance Techniques for Data Warehouses

Size: px
Start display at page:

Download "A Performance Analysis of View Maintenance Techniques for Data Warehouses"

Transcription

1 A Performance Analyss of Vew Mantenance Technques for Data Warehouses Xng Wang Dell Computer Corporaton Round Roc, Texas Le Gruenwald The nversty of Olahoma School of Computer Scence orman, OK 739 Guangtao Zhu nversty of Scence and Technology of Chna Department of Computer Scence and Technology Bejng, P.R. Chna 39 Abstract A data warehouse stores ntegrated nformaton as materalzed vews over data from one or more remote sources. These materalzed vews must be mantaned n response to actual relaton updates n the remote sources. Based on whether the current materalzed vews wll be used n computng the new vews, and whether the data warehouse wll query the remote data sources for addtonal data to do the computaton, the data warehouse vew mantenance technques are classfed nto four major categores: self-mantanable recomputaton, not self-mantanable recomputaton, self-mantanable ncremental mantenance, and not self-mantanable ncremental mantenance. Ths paper provdes a comprehensve comparson of the technques n these four categores n terms of the data warehouse space usage and number of rows accessed n order to propagate an update from a remote data source to a target materalzed vew n the data warehouse. The analyss shows that self-mantanable ncremental mantenance performs the best n terms of both space usage and number of rows accessed. Introducton A data warehouse stores ntegrated nformaton over data from one or more remote data sources for query and analyss [Hammer 95, Wener 96]. The ntegrated nformaton at the data warehouse s stored as materalzed vews. A vew s a vrtual relaton defned usng the actual relatons stored n the database. A materalzed vew s the result relaton of the evaluaton of the relatonal algebra expresson that defnes the vew relaton [Slberschatz 97]. sng these

2 materalzed vews, user queres can be answered qucly as the nformaton may be drectly avalable or can be calculated usng these materalzed vews on fly. A problem nown as the vew mantenance problem s how to mantan the materalzed vews so that they can be ept up to date n response to updates of the actual relatons n the remote data sources. There are numerous algorthms developed to solve the vew mantenance problem for tradtonal database systems. In these database systems, query expressons defned vews and actual relatons are stored at the same database. The database systems understand vew management and vew defntons and now what data s needed for propagatng updates to the vews. In a data warehouse, the query expressons that defne vews and actual relatons may be stored at dfferent database sources resdng at many stes. The sources may nform the data warehouse when an update occurs but they mght not be able to determne what data s needed for updatng the vews at the data warehouse. Therefore they may send only the actual data updates or the entre updated relatons to the data warehouse. pon recevng ths nformaton, the data warehouse may fnd that t needs some addtonal source data n order to update the vews. Then t wll ssue some queres to some of the sources to request the addtonal source data. Some of the sources may have updated ther data agan before they evaluate the requestng queres from the data warehouse. Therefore they wll send ncorrect addtonal data to the data warehouse, whch subsequently wll use the ncorrect data to compute the vews. Ths phenomenon s called dstrbuted vew mantenance anomaly. Solvng the vew mantenance problem n data warehouses s thus more complcated than that n tradtonal database systems. The objectves of ths paper are to provde a classfcaton of dfferent vew mantenance technques that have been proposed n the lterature and to conduct a comprehensve comparson

3 of these technques n terms of space usage and number of rows accessed usng the TPC benchmar for decson support queres. The rest of ths paper s organzed as follows. Secton classfes the exstng vew mantenance technques. Secton 3 presents a performance analyss of the technques. Fnally, Secton 4 provdes conclusons and future research. Classfcaton of Data Warehouse Vew Mantenance Technques Dependng on whether the current materalzed vews n a data warehouse are used n the computaton of ther new vews n response to updates that occur on the remote data sources, the exstng data warehouse vew mantenance technques can be classfed nto two broad categores: recomputaton and ncremental vew mantenance. Dependng on whether the data warehouse has to query the remote data sources n order to calculate the new vews, the technques can be further classfed as self-mantanable or not self-mantanable. The below subsectons dscuss these four categores.. The Self-Mantanable Recomputaton Category Materalzed vews can be computed from scratch by usng the vew defntons and other materalzed vews at the data warehouse. The current materalzed vews beng mantaned have no contrbuton to the calculaton of the new vews. Some technques replcate all or part of the remote data at the data warehouse. We can vew these replcated data as some nd of materalzed vews at the data warehouse. Others such as the self-mantenance warehouse approach dscussed n [Quass 96] store the remote relatons at the data warehouse as addtonal materalzed vews to provde data needed when the data warehouse computes the new vews. Therefore, the data warehouse wll never have to query the data sources for addtonal data. 3

4 A self-mantanable materalzed V vew can be defned n two ways. In the frst way, the materalzed vew V can be defned by usng other self-mantanable materalzed vews. In ths case, the vew V s defned as V Π proj cond v v v where all v 's are self-mantanable materalzed vews stored n the data warehouse. In the second way, vew V can be defned by usng the relatons resdng at the remote sources. In ths case, the vew defnton s V Π proj cond r r r where all r 's are self-mantanable relatons stored n the remote data sources. When there s an update occurrng at a remote data source, the data source nows what to do and wll send the update wth all related relatons n the vew defnton to the data warehouse n a transacton. Therefore, the data warehouse does not have to send a query requestng addtonal data to the remote data sources as all necessary nformaton s avalable. However, a self-mantanable data warehouse vew cannot be defned as V Π proj cond v v v r + r + r where all relatons v 's are self-mantanable materalzed vews resdng at the data warehouse, all relatons r 's are self-mantanable relatons resdng at the remote data sources, and there are totally relatons n the defnton of vew V. The reason s as follows. The data warehouse s stll self-mantanable when an update of relaton r l s propagated to the data warehouse as all v 's are avalable n the data warehouse. However, when an update of vew v s propagated to vew V, the relaton r l s not avalable at the data warehouse. Therefore, the data warehouse has to send 4

5 a query to the remote data sources to get the relaton r l n order to calculate vew V. Thus vew V s not self-mantanable. Although the above vew V s not self-mantanable, we can add some more selfmantanable materalzed vews and redefne vew V usng these self-mantanable vews to mae t self-mantanable. For example, n the above vew V where V Π proj cond v v v r + r + r suppose we can defne the followng self-mantanable vews v + Π proj cond r + r + r l and v + Π proj cond r l+ r l+ r Vew V can then be redefned as V Π proj cond v v v v + v + V s thus self-mantanable by defnton. The materalzed vew v at the data warehouse can be defned ether by other selfmantanable vews only, or by a collecton of self-mantanable relatons at the remote stes. We can vew the relatonshps among the data warehouse materalzed vews as a herarchy structure. The data warehouse may consst of many levels of materalzed vews. At the bottom level, there are materalzed vews defned by remote self-mantanable relatons. The vews defned by other self-mantanable materalzed vews are stayng on top of those defned by remote relatons. The materalzed vews n the data warehouse have to be mantaned n a correct order from the bottom level to the top level n the vew herarchy. The bottom level vews have to be mantaned frst. Then those vews n the second to the bottom level must be mantaned next, and so on. Fnally, the vews at the top of the herarchy are mantaned. Ths s the order that updates from the remote data sources are propagated to the data warehouse. 5

6 An advantage of the technques n ths category s that the vew mantenance anomaly problem s avoded as all necessary data are avalable at the data warehouse. The data warehouse nows the vew defntons and what to do wth the vews to eep them up to date. It elmnates accesses to the remote relatons, and therefore, t does not compete wth the remote data sources' local resources. The data warehouse mantenance operatons can then be totally separated from other OLTP operatons. Whether a remote data source s avalable or not wll not affect the data warehouse vew mantenance process. However, n order to mae the materalzed vews selfmantanable, addtonal materalzed vews that provde nformaton necessary for vew updates must be stored. Extra storage and tme are thus needed to mantan these addtonal vews.. The ot Self-Mantanable Recomputaton Category The data warehouse can recompute the materalzed vews from scratch usng the vew defntons, possbly some other materalzed vews at the data warehouse, and actual source relatons perodcally or whenever the source data s updated. That s, when an update occurs at the data source or perodcally, the source wll nform the data warehouse. Accordng to the query expresson that defnes the vew, the data warehouse may get part of data t wants from other materalzed vews at the data warehouse, and ssue queres to the sources to get the other data t does not have. The sources send the query results bac to the data warehouse. Based on the query results, the data warehouse calculates the vews and stores the results as materalzed vews n the data warehouse. The current materalzed vews have no contrbuton to the calculaton of the new vews. The data warehouse may replcate part of the remote relatons n the data warehouse. However, these data are not enough for mantanng the materalzed vews. Therefore, the data warehouse 6

7 wll have to query the remote data sources for addtonal data n order to mantan the vews. An extreme case s where the data warehouse does not replcate any remote relatons. If the vew mantenance process s not desgned carefully, the dstrbuted vew mantenance anomaly problem wll occur. Suppose that there s a data warehouse system where the remote data sources send updated relatons to the data warehouse whenever an update occurs at the data sources. pon recevng the nformaton, the data warehouse s ready to compute the new vews. But now let us assume that the data warehouse fnds that t needs some other relatons at some remote data sources to compute the new vews. It wll ssue queres to these data sources. Suppose the data sources that sent the updated relatons to the data warehouse update the relatons agan before they receve the queres from the data warehouse. The data sources answer the query and send the results to the data warehouse. These results mght contan extra nformaton that s ncorrect. The data warehouse wll then use the ncorrect data to compute the new vews, whch wll result n ncorrect new vews. There are a lot of solutons to the dstrbuted vew mantenance anomaly problem. The smplest soluton s as follows. Instead of sendng the updated relatons to the data warehouse, the data sources just smply nform the warehouse that there s an update that occurs at the data sources. The data warehouse wll ssue queres to the data sources to request all relatons requred n the vew defntons. After the data sources receve the queres, they wll send all requested relatons to the data warehouse. The anomaly problem s thus avoded. The not self-mantanable recomputaton approach s smple. The anomaly problem can be avoded easly. However, the recomputaton process s also tme and resource consumng. The data warehouse sends queres bac to the sources and wats for answers n order to computer 7

8 the new vews. Processng these queres consumes the sources local resources. If the sources are unavalable, the data warehouse wll not get the answers t needs..3 The Self-Mantanable Incremental Mantenance Category In ths category, the data warehouse vews are mantaned by usng the vew defntons, the materalzed vews, and the vew updates. The data warehouse wll never query the remote data sources as the nformaton at the data warehouse s enough for mantanng the vews. The data warehouse computes the vew updates, then adds them to the materalzed vews. The process s ncremental. ormally, only necessary remote relatons, or vews of the remote relatons are stored at the data warehouse as materalzed vews. In the extreme case, all remote relatons can be replcated at the data warehouse. The self-mantanable warehouse approaches dscussed n [Cu 99], [Hull 96] and [Quass 96] belong to ths category. Let us dscuss how to mantan a vew V that s defned as V Π proj cond v v v where each v s a materalzed vew and s defned as ether v Π proj cond v v v where each v j s a vew defned by other auxlary materalzed vews, or v Π proj cond r r r where each r j s a base relaton. In the frst case, each vew v j s defned ether as v j Π proj cond v v v where each v j s a vew defned by other auxlary materalzed vews, or 8

9 v j Π proj cond r r r where each r j s a base relaton. Fnally, at the lowest level of the vew herarchy dscussed earler n ths paper, vew v Mj can only be defned by relatons at the remote data sources as follows: v Mj Π proj cond r M r M r M where each r Mj s a base relaton. The above vew V s thus defned by M levels of the materalzed vews n the vew herarchy. In the second case, the vew can only be defned by base relatons r j. Each vew v j can be defned by v 3j. Vews can be defned by those vews at one level lower than t. The lowest level vews v Mj can only be defned by relatons at the remote data sources. Suppose an update j occurs at a data source, where represents the vew level n the vew herarchy and j represents the relaton n the vew defnton where the update occurs. The data source sends the update j along wth the related base relatons r, r, and r except r j to the data warehouse. The data warehouse calculates the vew update v -j usng the update j and relatons r, r, and r except r j as follows: v -j Π proj cond r r j r Then the vew update v -j s added to the vew v -j to produce the new vew. The vew v -j s thus mantaned. The vew update v -j s propagated to the vews at one level hgher than t. In ths case, t s v -j. The vew v -j s defned by self-mantanable vews as follows: v -j Π proj cond r - r - v -j r - We can calculate the vew update usng the followng formula: v -j Π proj cond r - r - v -j r - Ths process s repeated untl the top vew V n the vew herarchy s updated. 9

10 All ntermedate materalzed vews can be vewed as auxlary vews. These auxlary vews are self-mantanable. The materalzed vew V s self-mantanable by usng the update nformaton and addtonal nformaton from the auxlary vews. The data warehouse vews, ncludng vews such as V and auxlary vews, can be mantaned startng wth those vews that do not depend on any other auxlary vews, worng up to the fnal orgnal vew V. Fgure. Vew Herarchy Example All related materalzed prmary vews, auxlary vews and base relatons can be drawn n a herarchy structure as shown n Fgure. All leaves n the herarchy structure are those materalzed vews defned by the base relatons. In ths example, V s the prmary materalzed vew. Vews v, v and v are materalzed auxlary vews defned by other materalzed auxlary vews. Vews v 3, v, v 3, v 4, v 3 and v 4 are materalzed auxlary vews defned by the base relatons. All relatons r j 's are the base relatons. The vews n the leaves should be mantaned frst. Suppose an update for r 33 occurs n the data source. Vew v 4 should be mantaned frst. Then vews v and vew v must be mantaned next n that order. Fnally, the prmary vew V s mantaned. The data warehouse never needs to query the remote data source to get addtonal data. The data warehouse mantenance operatons can be totally separated from other OLTP

11 operatons. Whether the remote data source s avalable or not wll not affect the data warehouse vew mantenance process. However, n order to mae the materalzed vews self-mantanable, the auxlary vews are stored n the data warehouse to provde the addtonal nformaton. Extra storage and tme overhead are therefore requred to mantan the auxlary vews themselves. How to desgn materalzed vews at the data warehouse so that only necessary nformaton are stored at the data warehouse s a major ssue [Quass 96, Huyn 96a, 96b, 97b, 97c]..4 The ot Self-Mantanable Incremental Mantenance Category Instead of recomputng every vew from scratch, only parts of the warehouse that change are computed. However, the data warehouse has to query the remote data sources whenever necessary because the nformaton at the data warehouse s not enough to mantan the vew. A number of exstng approaches fall under ths category. Among them are the unrestrcted base access [Zhuge 95, 96, 97a] and runtme warehouse self-mantenance [Huyn 97a]..4. nrestrcted Base Access In the nrestrcted Base Access approach [Zhuge 95, 96, 97a], the data warehouse accesses the actual relatons from the data sources whenever necessary n order to mantan the materalzed vews. There are many proposed algorthms that follow ths approach. The Eager Compensatng Algorthm ECA s the smplest among them. It s also the fastest algorthm that wll let the data warehouse reman n a consstent state [Zhuge 98a and Zhuge 98b]. In ths algorthm, compensatng queres are sent bac to the sources to offset the effects of concurrent updates [Zhuge 95]. They are used only when the next update occurs at the sources before the sources receve the data warehouse queres. The data warehouse eeps a temporary table called COLLECT to eep the ntermedate answers t receves from the data sources. It also eeps a set called nanswered Query Set

12 QS to eep trac of those queres t sent to the data sources but has not receved ther answers yet. Suppose an update occurs at a data source. The data source sends to the data warehouse. Suppose the data warehouse wants to update the materalzed vew that s defned as V Π proj cond r r r where all relatons r 's resde at a sngle data source. The data warehouse determnes the query for calculatng the delta vew as Q Π proj cond r r r - r + r It then creates a temporary COLLECT table and QS set for processng ths specfc query, and sets both the COLLECT table and QS to empty. The data warehouse wrtes the query Q to the QS and sends the query Q to the data source. Suppose there s another update j that occurs at the same data source. The data source sends the update j to the data warehouse before t receves the query Q. The data warehouse now receves the update j. It nows that the upcomng answer for Q from the data source wll contan extra nformaton caused by smultaneous j update at the data source. Therefore, t has to offset ths extra nformaton. The data warehouse determnes the query Q j for updatng j as follows Q j Π proj cond r r r - r r + r j- j r j+ r Π proj cond r r r - r + r j- j r j+ r The frst part of Q j s the same as before; the second part s used to compensate for the extra nformaton that query Q wll receve. The data warehouse s eager to compensate before t receves the answer for query Q. Then t wrtes query Q j to QS and sends query Q j to the data source. ow the QS set contans two entres,.e., Q and Q j. The data source then processes

13 query Q and sends the answer to the data warehouse. The query answer A wll contan the followng extra nformaton due to the update j, Π proj cond r r r - r + r j- j r j+ r The data warehouse receves the answer A, removes entry Q from QS, wrtes the answer for Q to the temporary table COLLECT. ow QS contans only one entry Q j. The COLLECT table thus contans the followng data: A Π proj cond r r r - r + r j- r j r j+ r + Π proj cond r r r - r + r j- j r j+ r The data warehouse could not ntegrate the query answer nto the data warehouse at ths tme as t wll cause the data warehouse to contan nconsstent data. Therefore t has to wat for the query answer for query Q j to come bac. The COLLECT table s used to temporarly store query answers. After t receves the answer for the last unanswered query n QS n our case, Q j, t can ntegrate the data n the COLLECT table nto the data warehouse. The data source processes query Q j and sends answer A j to the data warehouse. A j contans the followng data: A j Π proj cond r r r - r r + r j- j r j+ r Π proj cond r r r - r + r j- j r j+ r The data warehouse receves the answer and removes the entry Q j from the QS set. ow QS s empty. The answer A j s ntegrated wth the data n the COLLECT table as follows: Π proj cond r r r - r + r j- r j r j+ r + Π proj cond r r r - r + r j- j r j+ r.+ Π proj cond r r r - r r + r j- j r j+ r Π proj cond r r r - r + r j- j r j+ r Π proj cond r r r - r + r j- r j r j+ r + Π proj cond r r r - r r + r j- j r j+ r 3

14 ow the COLLECT table contans the correct answer that the data warehouse wants. The data warehouse can then ntegrate the data n the COLLECT table nto the data warehouse. The nrestrcted Base Access approach s smple. Instead of dscardng old materalzed vews and calculatng new vews from scratch, ths approach calculates vew updates then adds them to the old vews n order to get the new vews. However, the data warehouse has to access remote source actual relatons n order to propagate one source update. In the eager compensatng algorthm dscussed above, all actual relatons have to be accessed f compensatng queres are used. In ths approach, the data warehouse may have to send queres bac to the sources and wats for answers n order to compute the vew updates. Therefore, ths approach has the same lmtaton as the not self-mantanable recomputaton approach. Computng these queres consumes remote sources local resources, and wll slow down other OLTP operatons. If the remote sources are unavalable, the data warehouse wll not get the answers t needs..4. Runtme Warehouse Self-Mantenance In the self-mantenance warehouse approach, at desgn tme, we determne vews to be selfmantanable and decde to add more auxlary vews for addtonal nformaton. Desgn-tme self-mantanablty s not flexble. It may be dffcult or mpossble for us to now the exact contents of the vews and ther updates at desgn tme. To solve ths problem, a run tme warehouse mantenance approach has been ntroduced [Huyn 97a]. The basc dea of the runtme self-mantenance approach s that the data warehouse generates the self-mantanable test for the vews to determne whether the vews are selfmantanable for a partcular update. At run tme, the self-mantanable test determnes the vews for self-mantanablty. If a vew s self-mantanable, then the vew can be mantaned by the 4

15 update nformaton, the vew tself, and the query expresson that defnes the vew. In ths case, the run-tme self-mantanable approach s smlar to the self-mantanable warehouse approach dscussed n the prevous secton. However, the data warehouse does not store and mantan any auxlary vews. If the vew s not self-mantanable, then the data warehouse has to query the remote data sources for those relatons t needs n order to update the vew. In ths case, ths approach s smlar to the unrestrcted base access approach. The runtme warehouse self-mantenance approach uses the vew mantanablty test frst before dong any mantenance. Ths creates the overhead n the data warehouse. However t mantans the vews by usng the vews n the data warehouse or some base relatons only, wthout requrng auxlary vews that are necessary n the self-mantanable ncremental mantenance approach. Ths saves the storage space n the data warehouse. 3 Performance Analyss We conduct an analyss to compare the performance of dfferent algorthms n the four categores. We consder only the problem of sngle vew mantenance n a sngle source envronment because the ECA algorthm n the not self-mantanable ncremental mantenance category can only be used n ths envronment [Zhuge 97a, Zhuge 96]. For self-mantanable vew mantenance technques, we also consder the case where all actual relatons are replcated at the data warehouse. 3. Performance Measurements In our analyss, only Select-Project-Jon vews are consdered. We measure the performances of the technques n terms of space and number of row accesses, whch are defned as follows: 5

16 !" Space: total space needed to store the data n the data warehouse, ncludng space for auxlary vews. We do not consder ndces.!" umber of rows accessed: the number of rows that must be accessed n the data warehouse and the data sources n order to ntegrate the updates nto the data warehouse. 3. Analyss Parameters The analyss parameters and ther default values are lsted n Table. The default values are calculated based on the TPC benchmar for decson support queres [TPC 99]. Meanng Symbol Default value Range nalty of vew V V 94 ~, Tuple sze of vew V n bytes tsv 43 ~ 5 umber of auxlary vews per vew av 3 ~ umber of base relatons n the vew 3 ~ 7 defnton nalty of base relaton r r 8,, ~,,, Tuple sze of base relaton r n tsr 6 ~ 8 bytes Selectvty: the fracton of tuples that.3. ~. satsfy the select condton The jon factor s roughly the fracton j.73. ~. of tuples n a relaton that jon wth others umber of tuples n a relaton that J j Calculated Calculated jon wth others r umber of nterferng updates for I.5 ~ each query nalty of update umber of tuple nsertons n a base relaton at data source update Table. Analyss Parameters 6

17 3.3 Comparson Based on Space eeded n the Data Warehouse 3.3. Self-Mantanable Recomputaton The technques n ths category do not query the remote data source for addtonal data n order to mantan the data warehouse materalzed vews. The data warehouse can replcate all or part of the remote base relatons at the data warehouse. These addtonal data tae space at the data warehouse. Here we consder the case where the materalzed vews are defned by other materalzed vews auxlary vews at the data warehouse, and all auxlary vews are replcated remote relatons. A vew V s defned as V Π proj cond v v v where all v 's are materalzed vews stored n the data warehouse. In the average case, the amount of space needed s: V ts V + av AV ts AV where av. If all auxlary vews are the same, the above formula can be reduced to V ts V + av AV ts AV av when none of the auxlary vews s needed. It means that the materalzed vew s a replcated remote relaton. Ths s the best case. The formula can be rewrtten as follows: V tsv If all auxlary vews are needed to mantan the vew, then av. Ths s the worst case. The formula can be rewrtten as follows: V ts V + AV ts AV If all auxlary vews are the same, the above formula can be reduced to V ts V + AV ts AV 7

18 3.3. ot Self-Mantanable Recomputaton Here we consder only the case where the data warehouse does not replcate any base relatons. Therefore, the data warehouse always has to query the remote data sources. The data warehouse stores only materalzed vews. In ths extreme stuaton, the amounts of space needed n the best case, the average case and the worst case are the same, and are equal to V tsv Self-Mantanable Incremental Mantenance Smlar to the self-mantanable recomputaton technques, the technques n ths category can replcate all or part of the remote data at the data warehouse. Here we consder only the case where the materalzed vews are defned by other materalzed vews auxlary vews at the data warehouse, and all auxlary vews are replcated remote relatons. A vew V s defned as V Π proj cond v v v where all v 's are materalzed vews stored n the data warehouse. The amount of space needed at the data warehouse s the same as that of the self-mantanable recomputaton technques. In the average case, the space needed s as follows: V ts V + av AV ts AV where av. If all auxlary vews are the same, the above formula can be reduced to V ts V + av AV ts AV In the best case, no auxlary vews are needed as the vew s self-mantanable. The formula can be rewrtten as follows: V tsv In the worst case, all auxlary vews are needed to mantan the vew. The formula can be rewrtten as follows: 8

19 V ts V + AV ts AV In the worst case, the auxlary vew's tuple sze and cardnalty are equal to those of the remote relaton. If all auxlary vews are the same, the above formula can be reduced to the followng: V ts V + AV ts AV ot Self-Mantanable Incremental Mantenance Here we consder the Eager Compensatng Algorthm ECA for ths category. In ECA, a temporary table COLLECT s used to store ntermedate query answers. For every update, the queres ncludng compensated queres are sent to the data source. ote that the COLLECT table s empty only when there s no query to the data source, or the answers for all the queres are returned to the data warehouse before a new update occurs at the data source. Ths s the best case. The space needed s as follows: V tsv In the average case, the sze of the COLLECT table for a specfc update wth all ts nterferng updates I s equal to the sum of the cardnaltes of the fnal query answers for all queres. The total number of queres sent to the data source can be calculated as below [Zhuge 97b]: q I j J In the worst case, the number of queres sent to the data source s: q. I r Let us derve the formula for the multple tuple update case. We follow the method used n [Zhuge 97b]. At frst, we have to fnd the number of wrapper queres sent from a vew 9

20 manager to the query processor [Zhuge 95, Zhuge 97b]. In the sngle tuple update case, A. In the multple tuple update case, A. The total number of wrapper queres to evaluate the update query s > : : : J q Assume all the s and Js are the same, the formula s smplfed to the followng: > : : J q The compensatng query s defned as: CQ Π proj cond n R n+ R n+ R There are n multple tuple updates and -n actual base relatons. The number of wrapper queres to evaluate the query s: > : : J j n q n n n n Then, we estmate the number of compensatng queres as I, where I s defned as the maxmum number of nterferng updates that occur between the tme when the data warehouse sends a query to the remote data source and the tme when the data warehouse receves the correspondng answer [Zhuge 98b]. In the worst case, there are I nterferng updates for each query. The number of wrapper queres s:

21 J j I q When, the above formula s reduced to the followng: J j I q whch s the formula of the sngle tuple update case [Zhuge 98b]. Suppose the cardnalty of the fnal query answer A for all nterferng updates are equal. J A for the sngle tuple update case [Zhuge 98b]. For the multple tuple update case, J A. When all s and Js are equal, J A. For a query wth n update relatons and n base relatons, n n n J j A, where n. The max cardnalty of the COLLECT table for updates at a tme s: J J j j I + J J j I. The space needed s V ts J J j I V ts V + +. In the worst case, when, j, J j r r. The space needed at the data warehouse s V ts r r I V ts V +.

22 3.3.5 Comparson Results In summary, for a gven vew V, the amounts of space needed n the data warehouse for the four categores are lsted n Table 3, where AV stands for Auxlary Vew. Category Self-Mantanable Recomputaton SMR ot Self-Mantanable Recomputaton SMR Self-Mantanable Incremental Mantenance SMIM ot Self-Mantanable Incremental Mantenance SIM Space needed n the data warehouse The best case: V tsv The average case: V ts V + av AV ts AV av The worst case: V ts V + AV ts AV V tsv The best case: V tsv The average case: V ts V + av AV ts AV av The worst case: V ts V + AV ts AV The best case: V tsv The average case: V ts V + + I j The worst case: V ts V + I r J J ts V r ts V Table. Space needed n the data warehouse We draw the graphs for the formulas derved above usng the parameters' defaults values lsted n Table. Due to space lmtaton, we wll only dscuss the results of the average case. In ths case, the not self-mantanable recomputaton approach does not requre extra space at the data warehouse. However, for the self-mantanable recomputaton and self-mantanable ncremental approaches, extra space s requred at the data warehouse to store the replcated

23 source data as auxlary vews. Ths extra space s proportonal to the total number of auxlary vews av, the auxlary vew cardnalty AV and the auxlary vew tuple sze tsav. For SIM, extra space s requred to store the COLLECT table. Suppose that the tuple sze of the COLLECT table s equal to the vew's tuple sze. In some cases, the space used to store the COLLECT table n SIM s small compared to the space used for storng the materalzed vews n SMR and SMIM. In other cases, the space used for storng the COLLECT table can be grown much larger than that of materalzed vews. Let us examne Fgures -4, where SMR s not shown snce t overlaps wth the x-axs. At frst, the space used for storng the COLLECT table s smaller and growng slower than that used for storng the materalzed auxlary vews when the cardnalty of auxlary vews s small Fgure. Then, the space used for storng the COLLECT table grows faster. At some pont AV 6. 7 n our case, the space used for storng the COLLECT table exceeds the space used for storng the materalzed auxlary vews Fgure 3. When the cardnaltes of auxlary vews or base relatons are large, the space used for storng the COLLECT table s much larger than that used to store the auxlary vews Fgure 4. From Table, we can see that the space needed for the SIM approach grows exponentally wth the select factor, jon factor, and cardnalty of updates. In summary, n the average case, as no addtonal space s requred at the data warehouse for the not self-mantanable approach, t s the best approach n terms of space used n the data warehouse. The amounts of space used for both the self-mantanable approaches, SMR and SMIR, are the same. When there are very few conflctng updates and the cardnaltes of the base relatons are relatvely small, the space used for the not self-mantanable approach s less than those of both the self-mantanable approaches. However, f there are a lot of conflctng 3

24 updates and the cardnaltes of the base relatons are large, the space used for the not selfmantanable approaches are much larger than those of both the self-mantanable approaches. Fgure. Space used n the data warehouse The average case Fgure 3. Space used n the data warehouse The average case 4

25 Fgure 4. Space used n the data warehouse The average case 3.4 Comparson Based on the umber of Rows Accessed To analyze the number of rows accessed at the data warehouse by the technques, we made the followng assumptons:!" The set of a prmary vew and ts auxlary vews f any s ndependent to other sets of prmary vews and ther auxlary vews.!" We do not consder ndces. Lnear search s thus used to chec f a record satsfes a select or jon condton.!" All auxlary vews are self-mantanable and are replcated base relatons.!" pdates to auxlary vews and prmary vews are for appendng only. 5

26 3.4. Self-Mantanable Recomputaton The data warehouse wll never query the remote data sources as all necessary data are avalable at the data warehouse. pdates from the remote data sources have to be propagated to the replcated relatons at the data warehouse frst, then the data warehouse recalculates the vew relaton and stores the result at the data warehouse as the new materalzed vew. In order to propagate an update to data warehouse replcated relaton, the number of rows to be accessed at the data warehouse s the cardnalty of the relaton tself plus the cardnalty of the update. That s, r +. Then the data warehouse recomputes the materalzed vew usng the vew defnton. Suppose the data warehouse materalzed vew s defned as: Q Π proj cond v v v The strategy such as the one descrbed n [Zhuge 97b] can be used to evaluate the query Q that defnes the vew V. Let us rename the actual relatons accordng to the jon order. The query Q that defnes the vew becomes Q Π proj cond R R R sng the equvalence rules [Slberschatz 97], the select condton cond can be decomposed nto a sequence of ndvdual selectons 's on ndvdual actual relatons R 's. Let T be the temporary relaton after s appled to R, and let T be the temporary relaton after the jon and selecton wth relaton R s done. T s the new vew. To compute T, each tuple n R has to be accessed for T - tmes. T T - R. T J T - and T 6

27 R. Each tuple n R s accessed once. R s accessed T R tmes. The number of tuples accessed for R and R s Thus, R + R R T J T J R. To get the new vew, the total number of tuples that must be accessed s R + R R + R 3 R J + + R R Π - J R + R + R 3 J + + R Π - J Suppose all the select and jon condtons, and cardnaltes of all base relatons are the same, the total number of tuples to be accessed n order to recompute the vew s equal to AV + AV + J + 3 J J - + AV AV J Ths s the average case. In the worst case, when, j, J j AV AV, the total number of tuples to be accessed s equal to AV + AV + AV + AV + + AV - AV + AV + AV + + AV - AV AV For update updates, the total number of tuples to be accessed n the average case s + update AV AV In the worst case, the formula s update AV AV J 7

28 Includng the number of rows that need to be accessed to mantan the vew and replcated data at the data warehouse, the total number of rows to be accessed n the average case s update V AV AV In the worst case, the number becomes update V AV 3.4. ot Self-Mantanable Recomputaton AV J Only source data s requred to be accessed. The reason s that the warehouse recalculates the full vew usng the source data each tme. It does not use the data warehouse data. Suppose the system locs all base relatons n order to evaluate the query expresson that defnes the vew. If the nested-loop jon method [Slberschatz 97] s used to evaluate t, the total number of rows to be accessed s r. Another strategy such as the one descrbed n [Zhuge 97b] can also be used to evaluate the query Q that defnes the vew V. It wll reduce the total number of rows to be accessed. Let us rename the actual relatons accordng to the jon order. The query Q that defnes the vew becomes Q Π proj cond R R R sng the same method presented n Secton 3.4., suppose all select and jon condtons, and cardnaltes of all base relatons are the same, the total number of tuples to be accessed n order to recompute the vew s equal to r + r + J + 3 J J - 8

29 + r r J Ths s the average case. In the worst case, when, j, J j r r, the total number of tuples to be accessed s equal to r + r + r + r + + r - r + r + r + + r - r r For update updates, the total number of tuples to be accessed n the average case s J + update r r. In the worst case, the formula s update r r Self-Mantanable Incremental Mantenance o queres are sent to the data sources for addtonal nformaton. Therefore, the number of rows accessed n the data source s equal to. For base relatons n a vew, av should be less than or equal to. In the worst case, av s equal to. At frst, the auxlary vew tself has to be mantaned before the prmary materalzed vew can be mantaned. Let stand for the cardnalty of update. Accordng to our assumpton that auxlary vews are self-mantanable and updates are used for appendng only, the number of rows needed to be accessed n order to mantan the auxlary vew s + AV. Let AV stand for the cardnalty of the auxlary vew update, whch s the same as. Then the update s propagated to the prmary vew. We need to calculate the 9

30 prmary vew update. The number of rows accessed can be estmated usng the same method presented n Secton Suppose the all select and jon condtons, and cardnaltes of all auxlary vews are the same, the total number of tuples to be accessed n order to calculate the prmary vew update s equal to AV + AV AV + J + 3 J + + av- J av- av + + AV AV J Fnally, the vew update s appended to the prmary vew. The total number of rows to be accessed n order to propagate a source update to the data warehouse s av V AV AV AV J Snce AV, we have av V AV AV AV J Ths s the average case. In the worst case, when, j, J j AV AV, av, the total number of tuples to be accessed s equal to V AV AV AV AV V AV AV AV In the sngle tuple update case,.e., AV, the total number of tuples to be accessed s equal to V AV AV For update updates, the average number of tuples to be accessed for the multple tuple update case s 3

31 av update V AV AV AV For the worst case, the total number of tuples to be accessed s update V AV AV ot Self-Mantanable Incremental Mantenance AV + J In the ECA algorthm, all tuples n the vew table have to be accessed n order to fnd a tuple to ntegrate wth the vew update. However, the data warehouse may have to access data from remote stes except for the best case. Parts of these queres are compensated. In Secton 3.3.4, we derve the number of wrapper queres correspondng to queres wth n relatons n the multple tuple update case as q I j J When there are totally update updates, the number of wrapper queres s q update I j J In general, for a query that s defned as Q Π proj cond R R R, the total number of tuples to be accessed n order to process the query s R + R + R 3 J + + R Π - J For a compensatng query Q Π proj cond R + R + R, there are multple tuple updates and - base relatons. Suppose all updates are evaluated frst, the number of rows to be accessed n order to evaluate a query s + + j j - + j - r Π - J 3

32 3 + + j j j - r + + j - r Π - J J r j j For the multple tuple update case, the total number of tuples to be accessed n order to propagate update updates from the remote source to the target data warehouse vew s: _ query per accessed tuples queres of number I update J r j j J j I update In the worst case, the total number s r r r I update Comparson Results The formulas to calculate the number of rows that need to be accessed n order to mantan a materalze vew for all four categores are lsted n Tables 3 and 4. We use the parameters' default values lsted n Table to draw the graphs showng the number of rows that need to be accessed to mantan a materalzed vew. Due to space lmtaton, here we wll dscuss the results for the average case only.

33 Category umber of rows accessed n data warehouse SMR The average case: update V + AV + + AV J umber of rows accessed n data source SMR SMIM The worst case: update V + + AV + update V AV The average case: update V + AV + AV + J av + AV av The worst case: update V + AV + AV + AV The best case: The average case: update r + r The worst case: update r r J Table 3. umber of Rows Accessed Part of 33

34 Category umber of rows accessed n data umber of rows accessed n data source warehouse SMIM update / I V The best case: The average case: update I + j j J + + j r The worst case: update I + r r + + J + r Table 4. umber of Rows Accessed Part of In the average case, the total numbers of rows accessed for the not self-mantanable recomputaton SMR approach and the self-mantanable recomputaton SMR approach are very smlar. As the same base relatons are replcated at the data warehouse n the selfmantanable recomputaton approach, actually these data wll be accessed twce. The frst tme s for mantanng the auxlary vews and the second tme s for mantanng the target materalzed vews at the data warehouse. When the cardnalty of the base relatons/auxlary vews, r, s small, the number of rows accessed for the self-mantanable recomputaton approach s somewhat larger than that of the not self-mantanable recomputaton approach Fgure 5. When r becomes larger, the dfference s small enough to be neglected Fgure 6. 34

35 Fgure 5. umber of rows accessed n the data warehouse The average case Fgure 6. umber of rows accessed n the data warehouse The average case 35

36 Fgure 7. umber of rows accessed n the data warehouse The average case Fgure 8. umber of rows accessed n the data warehouse The average case 36

37 When r s small, the numbers of rows accessed for self-mantanable recomputaton and not self-mantanable recomputaton grow faster than those for selfmantanable ncremental mantenance and not self-mantanable ncremental mantenance Fgure 7. When r becomes very large, the number of rows accessed for both the recomputaton approaches are much larger that those of both the ncremental approaches Fgure 8. In order to see the dfferences, we redrew Fgure 8 usng a logarthmc scale n the y-axs; the results are shown n Fgure 9. From Fgure 9, we can see that the number of rows accessed n SMR s the same as that n SMR, but s hgher than that of SIM, whch s hgher than that of SMIM. We also observe the smlar results when varyng the values of the select factor and jon factor j. Fgure 9. umber of rows accessed n the data warehouse The average case 37

38 4. Conclusons and Future Research Category Advantage Dsadvantage Self-Mantanable Recomputaton -Data warehouse vew mantenance operatons are totally separated from OLTP operatons; -navalable source wll not bloc the data warehouse vew mantenance process; -Data are replcated at data warehouse; -eed extra data storage for replcate data; -Have to mplement and mantan data transfer processes to transfer data from sources to data warehouse; ot Self- Mantanable Recomputaton -Very smple to mplement; -o replcate data at the data warehouse; -o extra data storage for replcate data; -Do not have to mplement and mantan data transfer processes to transfer data from sources to data warehouse; -navalable source wll bloc the data warehouse vew mantenance process; -Evaluatng queres at the data sources consumes local resources; -Data warehouse vew mantenance operatons are not separated from OLTP operatons; Self-Mantanable Incremental Mantenance ot Self- Mantanable Incremental Mantenance -Data warehouse vew mantenance operatons are totally separated from OLTP operatons; -navalable source wll not bloc the data warehouse vew mantenance process; -In the worst case, the number of rows accessed to mantan a vew s the lowest; -o replcate data at the data warehouse; -o extra data storage for replcate data; -Do not have to mplement and mantan data transfer processes to transfer data from sources to data warehouse; -Data are replcated at data warehouse; -eed extra data storage for replcate data; -Have to mplement and mantan data transfer processes to transfer data from sources to data warehouse; -navalable source wll bloc the data warehouse vew mantenance process; -Evaluatng queres at the data sources consume local resources; -Data warehouse vew mantenance operatons are not separated from OLTP operatons; -Have to desgn the vew mantenance process carefully to avod the anomaly problem; -In the worst case the number of rows accessed s the hghest; -Performance s down-graded rapdly; -eed extra storage for ntermedate data COLLECT tables; Table 5. Advantages and dsadvantages of the vew mantenance technques All data warehouse vew mantenance technques can be classfed nto four major categores. They are self-mantanable recomputaton, not self-mantanable recomputaton, self- 38

39 mantanable ncremental mantenance, and not self-mantanable ncremental mantenance. Ther advantages and dsadvantages are lsted n Table 5. Both self-mantanable recomputaton and self-mantanable ncremental mantenance approaches totally separate the data warehouse vew mantenance operatons from the OLTP operatons. Therefore, the vew mantenance operatons wll not consume data sources local resources. These operatons only consume the data warehouse's resources. Even f the remote data sources are not avalable, the data warehouse vew mantenance process can contnue runnng. However, a part or all source data are replcated at the data warehouse to mae the data warehouse vew mantenance process self-mantanable. These replcated data tae space. Data transfer processes are mplemented to transfer data from the remote data sources to the data warehouse. Desgn, mplement and mantan these processes are tme-consumng. A lot of unnecessary data may be duplcated at the data warehouse. However, these are the approaches that probably many large companes have to tae f they want to separate ther data warehouse vew mantenance operatons from ther OLTP operatons. Both the not self-mantanable recomputaton and not self-mantanable ncremental mantenance approaches suffer from some common dsadvantages. As the remote data sources have to process queres from the data warehouse that consume ther lmted local resources, the OLTP system wll be slow. Once a data source s unavalable, the data source wll not be able to answer queres sent from the data warehouse n tme. It wll bloc the data warehouse vew mantenance process. The not self-mantanable ncremental mantenance approach has some addtonal dsadvantages. To avod the anomaly problem, the vew mantenance process must be desgned carefully. If a lot of updates happen at the data sources, the data warehouse may ssue many compensatng queres. It s very possble that the data warehouse may never get the fnal 39

40 query results. Both approaches also have some common advantages. As there s no replcate data stored at the data warehouse, no data transfer process has to be mplemented and mantaned. There s no extra space for storng replcate data. Both approaches are good for small to mdszed companes whose OLTP database systems are not too busy. Among all the four categores, self-mantanable ncremental mantenance s the best n terms of space used n the data warehouse and number of rows accessed n order to propagate an update to the target materalzed vew n the data warehouse. As the cost of data storage becomes ncreasngly low, ths s the best approach to mplement a data warehouse. For future wor, we plan to consder another mportant performance measurement called the vew refresh tme, whch s defned as the elapse tme from the tme the system receves a source update to the tme the update s reflected n the vew. We wll also address the problem of multple vew mantenance n a multple source envronment. References [Cu 99] Y. Cu and J. Wdom. "Storng Auxlary Data for Effcent Vew Mantenance and Lneage Tracng." Techncal Report, Stanford nversty, 999. [Hammer 95] J. Hammer, H. Garca-Molna, J. Wdom, W. Labo, and Y. Zhuge, The Stanford Data Warehousng Project. IEEE Data Engneerng Bulletn, June 995. [Huyn 96a]. Huyn, Effcent Vew Self-Mantenance. Proceedngs of the ACM Worshop on Materalzed Vews: Technques and Applcatons, Montreal, Canada, June 7, 996. [Huyn 96b]. Huyn, Effcent Self-Mantenance of Materalzed Vews. Techncal ote, 996. [Huyn 97a]. Huyn, Mantanng Data Warehouse nder Lmted Source Access. Ph.D. Thess, Stanford nversty, August 997. [Huyn 97b]. Huyn, Explotng Dependences to Enhance Vew Self-Mantanablty. Techncal ote, 997. [Huyn 97c]. Huyn, Multple-Vew Self-Mantenance n Data Warehousng Envronments. Proceedngs of the 3rd VLDB Conference, Athens, Greece, 997. [Hull 96] R. Hull and G. Zhou, A framewor for supportng data ntegraton usng the materalzed and vrtual approaches, In SIGMOD 996. [Labo 97b] W. Labo, D. Quass, and B. Adelberg, Physcal Database Desgn for Data Warehousng. Proceedngs of the Internatonal Conference on Data Engneerng, Bnghamton, K, Aprl,

Luby s Alg. for Maximal Independent Sets using Pairwise Independence

Luby s Alg. for Maximal Independent Sets using Pairwise Independence Lecture Notes for Randomzed Algorthms Luby s Alg. for Maxmal Independent Sets usng Parwse Independence Last Updated by Erc Vgoda on February, 006 8. Maxmal Independent Sets For a graph G = (V, E), an ndependent

More information

1. Fundamentals of probability theory 2. Emergence of communication traffic 3. Stochastic & Markovian Processes (SP & MP)

1. Fundamentals of probability theory 2. Emergence of communication traffic 3. Stochastic & Markovian Processes (SP & MP) 6.3 / -- Communcaton Networks II (Görg) SS20 -- www.comnets.un-bremen.de Communcaton Networks II Contents. Fundamentals of probablty theory 2. Emergence of communcaton traffc 3. Stochastc & Markovan Processes

More information

benefit is 2, paid if the policyholder dies within the year, and probability of death within the year is ).

benefit is 2, paid if the policyholder dies within the year, and probability of death within the year is ). REVIEW OF RISK MANAGEMENT CONCEPTS LOSS DISTRIBUTIONS AND INSURANCE Loss and nsurance: When someone s subject to the rsk of ncurrng a fnancal loss, the loss s generally modeled usng a random varable or

More information

A Secure Password-Authenticated Key Agreement Using Smart Cards

A Secure Password-Authenticated Key Agreement Using Smart Cards A Secure Password-Authentcated Key Agreement Usng Smart Cards Ka Chan 1, Wen-Chung Kuo 2 and Jn-Chou Cheng 3 1 Department of Computer and Informaton Scence, R.O.C. Mltary Academy, Kaohsung 83059, Tawan,

More information

"Research Note" APPLICATION OF CHARGE SIMULATION METHOD TO ELECTRIC FIELD CALCULATION IN THE POWER CABLES *

Research Note APPLICATION OF CHARGE SIMULATION METHOD TO ELECTRIC FIELD CALCULATION IN THE POWER CABLES * Iranan Journal of Scence & Technology, Transacton B, Engneerng, ol. 30, No. B6, 789-794 rnted n The Islamc Republc of Iran, 006 Shraz Unversty "Research Note" ALICATION OF CHARGE SIMULATION METHOD TO ELECTRIC

More information

Calculation of Sampling Weights

Calculation of Sampling Weights Perre Foy Statstcs Canada 4 Calculaton of Samplng Weghts 4.1 OVERVIEW The basc sample desgn used n TIMSS Populatons 1 and 2 was a two-stage stratfed cluster desgn. 1 The frst stage conssted of a sample

More information

Frequency Selective IQ Phase and IQ Amplitude Imbalance Adjustments for OFDM Direct Conversion Transmitters

Frequency Selective IQ Phase and IQ Amplitude Imbalance Adjustments for OFDM Direct Conversion Transmitters Frequency Selectve IQ Phase and IQ Ampltude Imbalance Adjustments for OFDM Drect Converson ransmtters Edmund Coersmeer, Ernst Zelnsk Noka, Meesmannstrasse 103, 44807 Bochum, Germany edmund.coersmeer@noka.com,

More information

DEFINING %COMPLETE IN MICROSOFT PROJECT

DEFINING %COMPLETE IN MICROSOFT PROJECT CelersSystems DEFINING %COMPLETE IN MICROSOFT PROJECT PREPARED BY James E Aksel, PMP, PMI-SP, MVP For Addtonal Informaton about Earned Value Management Systems and reportng, please contact: CelersSystems,

More information

An Alternative Way to Measure Private Equity Performance

An Alternative Way to Measure Private Equity Performance An Alternatve Way to Measure Prvate Equty Performance Peter Todd Parlux Investment Technology LLC Summary Internal Rate of Return (IRR) s probably the most common way to measure the performance of prvate

More information

Study on Model of Risks Assessment of Standard Operation in Rural Power Network

Study on Model of Risks Assessment of Standard Operation in Rural Power Network Study on Model of Rsks Assessment of Standard Operaton n Rural Power Network Qngj L 1, Tao Yang 2 1 Qngj L, College of Informaton and Electrcal Engneerng, Shenyang Agrculture Unversty, Shenyang 110866,

More information

What is Candidate Sampling

What is Candidate Sampling What s Canddate Samplng Say we have a multclass or mult label problem where each tranng example ( x, T ) conssts of a context x a small (mult)set of target classes T out of a large unverse L of possble

More information

Lecture 3: Force of Interest, Real Interest Rate, Annuity

Lecture 3: Force of Interest, Real Interest Rate, Annuity Lecture 3: Force of Interest, Real Interest Rate, Annuty Goals: Study contnuous compoundng and force of nterest Dscuss real nterest rate Learn annuty-mmedate, and ts present value Study annuty-due, and

More information

Single and multiple stage classifiers implementing logistic discrimination

Single and multiple stage classifiers implementing logistic discrimination Sngle and multple stage classfers mplementng logstc dscrmnaton Hélo Radke Bttencourt 1 Dens Alter de Olvera Moraes 2 Vctor Haertel 2 1 Pontfíca Unversdade Católca do Ro Grande do Sul - PUCRS Av. Ipranga,

More information

CHOLESTEROL REFERENCE METHOD LABORATORY NETWORK. Sample Stability Protocol

CHOLESTEROL REFERENCE METHOD LABORATORY NETWORK. Sample Stability Protocol CHOLESTEROL REFERENCE METHOD LABORATORY NETWORK Sample Stablty Protocol Background The Cholesterol Reference Method Laboratory Network (CRMLN) developed certfcaton protocols for total cholesterol, HDL

More information

8.5 UNITARY AND HERMITIAN MATRICES. The conjugate transpose of a complex matrix A, denoted by A*, is given by

8.5 UNITARY AND HERMITIAN MATRICES. The conjugate transpose of a complex matrix A, denoted by A*, is given by 6 CHAPTER 8 COMPLEX VECTOR SPACES 5. Fnd the kernel of the lnear transformaton gven n Exercse 5. In Exercses 55 and 56, fnd the mage of v, for the ndcated composton, where and are gven by the followng

More information

Time Value of Money Module

Time Value of Money Module Tme Value of Money Module O BJECTIVES After readng ths Module, you wll be able to: Understand smple nterest and compound nterest. 2 Compute and use the future value of a sngle sum. 3 Compute and use the

More information

The OC Curve of Attribute Acceptance Plans

The OC Curve of Attribute Acceptance Plans The OC Curve of Attrbute Acceptance Plans The Operatng Characterstc (OC) curve descrbes the probablty of acceptng a lot as a functon of the lot s qualty. Fgure 1 shows a typcal OC Curve. 10 8 6 4 1 3 4

More information

Project Networks With Mixed-Time Constraints

Project Networks With Mixed-Time Constraints Project Networs Wth Mxed-Tme Constrants L Caccetta and B Wattananon Western Australan Centre of Excellence n Industral Optmsaton (WACEIO) Curtn Unversty of Technology GPO Box U1987 Perth Western Australa

More information

Lecture 3: Annuity. Study annuities whose payments form a geometric progression or a arithmetic progression.

Lecture 3: Annuity. Study annuities whose payments form a geometric progression or a arithmetic progression. Lecture 3: Annuty Goals: Learn contnuous annuty and perpetuty. Study annutes whose payments form a geometrc progresson or a arthmetc progresson. Dscuss yeld rates. Introduce Amortzaton Suggested Textbook

More information

Performance attribution for multi-layered investment decisions

Performance attribution for multi-layered investment decisions Performance attrbuton for mult-layered nvestment decsons 880 Thrd Avenue 7th Floor Ne Yor, NY 10022 212.866.9200 t 212.866.9201 f qsnvestors.com Inna Oounova Head of Strategc Asset Allocaton Portfolo Management

More information

Section 5.4 Annuities, Present Value, and Amortization

Section 5.4 Annuities, Present Value, and Amortization Secton 5.4 Annutes, Present Value, and Amortzaton Present Value In Secton 5.2, we saw that the present value of A dollars at nterest rate per perod for n perods s the amount that must be deposted today

More information

Software project management with GAs

Software project management with GAs Informaton Scences 177 (27) 238 241 www.elsever.com/locate/ns Software project management wth GAs Enrque Alba *, J. Francsco Chcano Unversty of Málaga, Grupo GISUM, Departamento de Lenguajes y Cencas de

More information

Forecasting the Direction and Strength of Stock Market Movement

Forecasting the Direction and Strength of Stock Market Movement Forecastng the Drecton and Strength of Stock Market Movement Jngwe Chen Mng Chen Nan Ye cjngwe@stanford.edu mchen5@stanford.edu nanye@stanford.edu Abstract - Stock market s one of the most complcated systems

More information

Simple Interest Loans (Section 5.1) :

Simple Interest Loans (Section 5.1) : Chapter 5 Fnance The frst part of ths revew wll explan the dfferent nterest and nvestment equatons you learned n secton 5.1 through 5.4 of your textbook and go through several examples. The second part

More information

Research of concurrency control protocol based on the main memory database

Research of concurrency control protocol based on the main memory database Research of concurrency control protocol based on the man memory database Abstract Yonghua Zhang * Shjazhuang Unversty of economcs, Shjazhuang, Shjazhuang, Chna Receved 1 October 2014, www.cmnt.lv The

More information

Implementation of Deutsch's Algorithm Using Mathcad

Implementation of Deutsch's Algorithm Using Mathcad Implementaton of Deutsch's Algorthm Usng Mathcad Frank Roux The followng s a Mathcad mplementaton of Davd Deutsch's quantum computer prototype as presented on pages - n "Machnes, Logc and Quantum Physcs"

More information

Support Vector Machines

Support Vector Machines Support Vector Machnes Max Wellng Department of Computer Scence Unversty of Toronto 10 Kng s College Road Toronto, M5S 3G5 Canada wellng@cs.toronto.edu Abstract Ths s a note to explan support vector machnes.

More information

Hybrid Log-based Fault Tolerant scheme for Mobile Computing System Yongning Zhai1, a, Zhenpeng Xu1, b, Weini Zeng1, c

Hybrid Log-based Fault Tolerant scheme for Mobile Computing System Yongning Zhai1, a, Zhenpeng Xu1, b, Weini Zeng1, c 2nd Internatonal Conference on Advances n Mechancal Engneerng and Industral Informatcs (AMEII 2016) Hybrd Log-based Fault Tolerant scheme for Moble Computng System Yongnng Zha1, a, Zhenpeng Xu1, b, Wen

More information

On the Optimal Control of a Cascade of Hydro-Electric Power Stations

On the Optimal Control of a Cascade of Hydro-Electric Power Stations On the Optmal Control of a Cascade of Hydro-Electrc Power Statons M.C.M. Guedes a, A.F. Rbero a, G.V. Smrnov b and S. Vlela c a Department of Mathematcs, School of Scences, Unversty of Porto, Portugal;

More information

A Replication-Based and Fault Tolerant Allocation Algorithm for Cloud Computing

A Replication-Based and Fault Tolerant Allocation Algorithm for Cloud Computing A Replcaton-Based and Fault Tolerant Allocaton Algorthm for Cloud Computng Tork Altameem Dept of Computer Scence, RCC, Kng Saud Unversty, PO Box: 28095 11437 Ryadh-Saud Araba Abstract The very large nfrastructure

More information

Data Broadcast on a Multi-System Heterogeneous Overlayed Wireless Network *

Data Broadcast on a Multi-System Heterogeneous Overlayed Wireless Network * JOURNAL OF INFORMATION SCIENCE AND ENGINEERING 24, 819-840 (2008) Data Broadcast on a Mult-System Heterogeneous Overlayed Wreless Network * Department of Computer Scence Natonal Chao Tung Unversty Hsnchu,

More information

Recurrence. 1 Definitions and main statements

Recurrence. 1 Definitions and main statements Recurrence 1 Defntons and man statements Let X n, n = 0, 1, 2,... be a MC wth the state space S = (1, 2,...), transton probabltes p j = P {X n+1 = j X n = }, and the transton matrx P = (p j ),j S def.

More information

Section 5.3 Annuities, Future Value, and Sinking Funds

Section 5.3 Annuities, Future Value, and Sinking Funds Secton 5.3 Annutes, Future Value, and Snkng Funds Ordnary Annutes A sequence of equal payments made at equal perods of tme s called an annuty. The tme between payments s the payment perod, and the tme

More information

10.2 Future Value and Present Value of an Ordinary Simple Annuity

10.2 Future Value and Present Value of an Ordinary Simple Annuity 348 Chapter 10 Annutes 10.2 Future Value and Present Value of an Ordnary Smple Annuty In compound nterest, 'n' s the number of compoundng perods durng the term. In an ordnary smple annuty, payments are

More information

An Analysis of Central Processor Scheduling in Multiprogrammed Computer Systems

An Analysis of Central Processor Scheduling in Multiprogrammed Computer Systems STAN-CS-73-355 I SU-SE-73-013 An Analyss of Central Processor Schedulng n Multprogrammed Computer Systems (Dgest Edton) by Thomas G. Prce October 1972 Techncal Report No. 57 Reproducton n whole or n part

More information

Institute of Informatics, Faculty of Business and Management, Brno University of Technology,Czech Republic

Institute of Informatics, Faculty of Business and Management, Brno University of Technology,Czech Republic Lagrange Multplers as Quanttatve Indcators n Economcs Ivan Mezník Insttute of Informatcs, Faculty of Busness and Management, Brno Unversty of TechnologCzech Republc Abstract The quanttatve role of Lagrange

More information

Power-of-Two Policies for Single- Warehouse Multi-Retailer Inventory Systems with Order Frequency Discounts

Power-of-Two Policies for Single- Warehouse Multi-Retailer Inventory Systems with Order Frequency Discounts Power-of-wo Polces for Sngle- Warehouse Mult-Retaler Inventory Systems wth Order Frequency Dscounts José A. Ventura Pennsylvana State Unversty (USA) Yale. Herer echnon Israel Insttute of echnology (Israel)

More information

RESEARCH ON DUAL-SHAKER SINE VIBRATION CONTROL. Yaoqi FENG 1, Hanping QIU 1. China Academy of Space Technology (CAST) yaoqi.feng@yahoo.

RESEARCH ON DUAL-SHAKER SINE VIBRATION CONTROL. Yaoqi FENG 1, Hanping QIU 1. China Academy of Space Technology (CAST) yaoqi.feng@yahoo. ICSV4 Carns Australa 9- July, 007 RESEARCH ON DUAL-SHAKER SINE VIBRATION CONTROL Yaoq FENG, Hanpng QIU Dynamc Test Laboratory, BISEE Chna Academy of Space Technology (CAST) yaoq.feng@yahoo.com Abstract

More information

A Novel Methodology of Working Capital Management for Large. Public Constructions by Using Fuzzy S-curve Regression

A Novel Methodology of Working Capital Management for Large. Public Constructions by Using Fuzzy S-curve Regression Novel Methodology of Workng Captal Management for Large Publc Constructons by Usng Fuzzy S-curve Regresson Cheng-Wu Chen, Morrs H. L. Wang and Tng-Ya Hseh Department of Cvl Engneerng, Natonal Central Unversty,

More information

Extending Probabilistic Dynamic Epistemic Logic

Extending Probabilistic Dynamic Epistemic Logic Extendng Probablstc Dynamc Epstemc Logc Joshua Sack May 29, 2008 Probablty Space Defnton A probablty space s a tuple (S, A, µ), where 1 S s a set called the sample space. 2 A P(S) s a σ-algebra: a set

More information

Mining Multiple Large Data Sources

Mining Multiple Large Data Sources The Internatonal Arab Journal of Informaton Technology, Vol. 7, No. 3, July 2 24 Mnng Multple Large Data Sources Anmesh Adhkar, Pralhad Ramachandrarao 2, Bhanu Prasad 3, and Jhml Adhkar 4 Department of

More information

The Development of Web Log Mining Based on Improve-K-Means Clustering Analysis

The Development of Web Log Mining Based on Improve-K-Means Clustering Analysis The Development of Web Log Mnng Based on Improve-K-Means Clusterng Analyss TngZhong Wang * College of Informaton Technology, Luoyang Normal Unversty, Luoyang, 471022, Chna wangtngzhong2@sna.cn Abstract.

More information

FORMAL ANALYSIS FOR REAL-TIME SCHEDULING

FORMAL ANALYSIS FOR REAL-TIME SCHEDULING FORMAL ANALYSIS FOR REAL-TIME SCHEDULING Bruno Dutertre and Vctora Stavrdou, SRI Internatonal, Menlo Park, CA Introducton In modern avoncs archtectures, applcaton software ncreasngly reles on servces provded

More information

Period and Deadline Selection for Schedulability in Real-Time Systems

Period and Deadline Selection for Schedulability in Real-Time Systems Perod and Deadlne Selecton for Schedulablty n Real-Tme Systems Thdapat Chantem, Xaofeng Wang, M.D. Lemmon, and X. Sharon Hu Department of Computer Scence and Engneerng, Department of Electrcal Engneerng

More information

Conferencing protocols and Petri net analysis

Conferencing protocols and Petri net analysis Conferencng protocols and Petr net analyss E. ANTONIDAKIS Department of Electroncs, Technologcal Educatonal Insttute of Crete, GREECE ena@chana.tecrete.gr Abstract: Durng a computer conference, users desre

More information

Answer: A). There is a flatter IS curve in the high MPC economy. Original LM LM after increase in M. IS curve for low MPC economy

Answer: A). There is a flatter IS curve in the high MPC economy. Original LM LM after increase in M. IS curve for low MPC economy 4.02 Quz Solutons Fall 2004 Multple-Choce Questons (30/00 ponts) Please, crcle the correct answer for each of the followng 0 multple-choce questons. For each queston, only one of the answers s correct.

More information

Multiple-Period Attribution: Residuals and Compounding

Multiple-Period Attribution: Residuals and Compounding Multple-Perod Attrbuton: Resduals and Compoundng Our revewer gave these authors full marks for dealng wth an ssue that performance measurers and vendors often regard as propretary nformaton. In 1994, Dens

More information

Rate Monotonic (RM) Disadvantages of cyclic. TDDB47 Real Time Systems. Lecture 2: RM & EDF. Priority-based scheduling. States of a process

Rate Monotonic (RM) Disadvantages of cyclic. TDDB47 Real Time Systems. Lecture 2: RM & EDF. Priority-based scheduling. States of a process Dsadvantages of cyclc TDDB47 Real Tme Systems Manual scheduler constructon Cannot deal wth any runtme changes What happens f we add a task to the set? Real-Tme Systems Laboratory Department of Computer

More information

Financial Mathemetics

Financial Mathemetics Fnancal Mathemetcs 15 Mathematcs Grade 12 Teacher Gude Fnancal Maths Seres Overvew In ths seres we am to show how Mathematcs can be used to support personal fnancal decsons. In ths seres we jon Tebogo,

More information

RequIn, a tool for fast web traffic inference

RequIn, a tool for fast web traffic inference RequIn, a tool for fast web traffc nference Olver aul, Jean Etenne Kba GET/INT, LOR Department 9 rue Charles Fourer 90 Evry, France Olver.aul@nt-evry.fr, Jean-Etenne.Kba@nt-evry.fr Abstract As networked

More information

A Design Method of High-availability and Low-optical-loss Optical Aggregation Network Architecture

A Design Method of High-availability and Low-optical-loss Optical Aggregation Network Architecture A Desgn Method of Hgh-avalablty and Low-optcal-loss Optcal Aggregaton Network Archtecture Takehro Sato, Kuntaka Ashzawa, Kazumasa Tokuhash, Dasuke Ish, Satoru Okamoto and Naoak Yamanaka Dept. of Informaton

More information

A role based access in a hierarchical sensor network architecture to provide multilevel security

A role based access in a hierarchical sensor network architecture to provide multilevel security 1 A role based access n a herarchcal sensor network archtecture to provde multlevel securty Bswajt Panja a Sanjay Kumar Madra b and Bharat Bhargava c a Department of Computer Scenc Morehead State Unversty

More information

Fault tolerance in cloud technologies presented as a service

Fault tolerance in cloud technologies presented as a service Internatonal Scentfc Conference Computer Scence 2015 Pavel Dzhunev, PhD student Fault tolerance n cloud technologes presented as a servce INTRODUCTION Improvements n technques for vrtualzaton and performance

More information

How To Understand The Results Of The German Meris Cloud And Water Vapour Product

How To Understand The Results Of The German Meris Cloud And Water Vapour Product Ttel: Project: Doc. No.: MERIS level 3 cloud and water vapour products MAPP MAPP-ATBD-ClWVL3 Issue: 1 Revson: 0 Date: 9.12.1998 Functon Name Organsaton Sgnature Date Author: Bennartz FUB Preusker FUB Schüller

More information

Research Article Enhanced Two-Step Method via Relaxed Order of α-satisfactory Degrees for Fuzzy Multiobjective Optimization

Research Article Enhanced Two-Step Method via Relaxed Order of α-satisfactory Degrees for Fuzzy Multiobjective Optimization Hndaw Publshng Corporaton Mathematcal Problems n Engneerng Artcle ID 867836 pages http://dxdoorg/055/204/867836 Research Artcle Enhanced Two-Step Method va Relaxed Order of α-satsfactory Degrees for Fuzzy

More information

Can Auto Liability Insurance Purchases Signal Risk Attitude?

Can Auto Liability Insurance Purchases Signal Risk Attitude? Internatonal Journal of Busness and Economcs, 2011, Vol. 10, No. 2, 159-164 Can Auto Lablty Insurance Purchases Sgnal Rsk Atttude? Chu-Shu L Department of Internatonal Busness, Asa Unversty, Tawan Sheng-Chang

More information

Joint Scheduling of Processing and Shuffle Phases in MapReduce Systems

Joint Scheduling of Processing and Shuffle Phases in MapReduce Systems Jont Schedulng of Processng and Shuffle Phases n MapReduce Systems Fangfe Chen, Mural Kodalam, T. V. Lakshman Department of Computer Scence and Engneerng, The Penn State Unversty Bell Laboratores, Alcatel-Lucent

More information

SUPPLIER FINANCING AND STOCK MANAGEMENT. A JOINT VIEW.

SUPPLIER FINANCING AND STOCK MANAGEMENT. A JOINT VIEW. SUPPLIER FINANCING AND STOCK MANAGEMENT. A JOINT VIEW. Lucía Isabel García Cebrán Departamento de Economía y Dreccón de Empresas Unversdad de Zaragoza Gran Vía, 2 50.005 Zaragoza (Span) Phone: 976-76-10-00

More information

Calculating the high frequency transmission line parameters of power cables

Calculating the high frequency transmission line parameters of power cables < ' Calculatng the hgh frequency transmsson lne parameters of power cables Authors: Dr. John Dcknson, Laboratory Servces Manager, N 0 RW E B Communcatons Mr. Peter J. Ncholson, Project Assgnment Manager,

More information

An Interest-Oriented Network Evolution Mechanism for Online Communities

An Interest-Oriented Network Evolution Mechanism for Online Communities An Interest-Orented Network Evoluton Mechansm for Onlne Communtes Cahong Sun and Xaopng Yang School of Informaton, Renmn Unversty of Chna, Bejng 100872, P.R. Chna {chsun,yang}@ruc.edu.cn Abstract. Onlne

More information

The Greedy Method. Introduction. 0/1 Knapsack Problem

The Greedy Method. Introduction. 0/1 Knapsack Problem The Greedy Method Introducton We have completed data structures. We now are gong to look at algorthm desgn methods. Often we are lookng at optmzaton problems whose performance s exponental. For an optmzaton

More information

Number of Levels Cumulative Annual operating Income per year construction costs costs ($) ($) ($) 1 600,000 35,000 100,000 2 2,200,000 60,000 350,000

Number of Levels Cumulative Annual operating Income per year construction costs costs ($) ($) ($) 1 600,000 35,000 100,000 2 2,200,000 60,000 350,000 Problem Set 5 Solutons 1 MIT s consderng buldng a new car park near Kendall Square. o unversty funds are avalable (overhead rates are under pressure and the new faclty would have to pay for tself from

More information

Module 2 LOSSLESS IMAGE COMPRESSION SYSTEMS. Version 2 ECE IIT, Kharagpur

Module 2 LOSSLESS IMAGE COMPRESSION SYSTEMS. Version 2 ECE IIT, Kharagpur Module LOSSLESS IMAGE COMPRESSION SYSTEMS Lesson 3 Lossless Compresson: Huffman Codng Instructonal Objectves At the end of ths lesson, the students should be able to:. Defne and measure source entropy..

More information

SPEE Recommended Evaluation Practice #6 Definition of Decline Curve Parameters Background:

SPEE Recommended Evaluation Practice #6 Definition of Decline Curve Parameters Background: SPEE Recommended Evaluaton Practce #6 efnton of eclne Curve Parameters Background: The producton hstores of ol and gas wells can be analyzed to estmate reserves and future ol and gas producton rates and

More information

Enabling P2P One-view Multi-party Video Conferencing

Enabling P2P One-view Multi-party Video Conferencing Enablng P2P One-vew Mult-party Vdeo Conferencng Yongxang Zhao, Yong Lu, Changja Chen, and JanYn Zhang Abstract Mult-Party Vdeo Conferencng (MPVC) facltates realtme group nteracton between users. Whle P2P

More information

Open Access A Load Balancing Strategy with Bandwidth Constraint in Cloud Computing. Jing Deng 1,*, Ping Guo 2, Qi Li 3, Haizhu Chen 1

Open Access A Load Balancing Strategy with Bandwidth Constraint in Cloud Computing. Jing Deng 1,*, Ping Guo 2, Qi Li 3, Haizhu Chen 1 Send Orders for Reprnts to reprnts@benthamscence.ae The Open Cybernetcs & Systemcs Journal, 2014, 8, 115-121 115 Open Access A Load Balancng Strategy wth Bandwdth Constrant n Cloud Computng Jng Deng 1,*,

More information

An MILP model for planning of batch plants operating in a campaign-mode

An MILP model for planning of batch plants operating in a campaign-mode An MILP model for plannng of batch plants operatng n a campagn-mode Yanna Fumero Insttuto de Desarrollo y Dseño CONICET UTN yfumero@santafe-concet.gov.ar Gabrela Corsano Insttuto de Desarrollo y Dseño

More information

ANALYZING THE RELATIONSHIPS BETWEEN QUALITY, TIME, AND COST IN PROJECT MANAGEMENT DECISION MAKING

ANALYZING THE RELATIONSHIPS BETWEEN QUALITY, TIME, AND COST IN PROJECT MANAGEMENT DECISION MAKING ANALYZING THE RELATIONSHIPS BETWEEN QUALITY, TIME, AND COST IN PROJECT MANAGEMENT DECISION MAKING Matthew J. Lberatore, Department of Management and Operatons, Vllanova Unversty, Vllanova, PA 19085, 610-519-4390,

More information

To manage leave, meeting institutional requirements and treating individual staff members fairly and consistently.

To manage leave, meeting institutional requirements and treating individual staff members fairly and consistently. Corporate Polces & Procedures Human Resources - Document CPP216 Leave Management Frst Produced: Current Verson: Past Revsons: Revew Cycle: Apples From: 09/09/09 26/10/12 09/09/09 3 years Immedately Authorsaton:

More information

FREQUENCY OF OCCURRENCE OF CERTAIN CHEMICAL CLASSES OF GSR FROM VARIOUS AMMUNITION TYPES

FREQUENCY OF OCCURRENCE OF CERTAIN CHEMICAL CLASSES OF GSR FROM VARIOUS AMMUNITION TYPES FREQUENCY OF OCCURRENCE OF CERTAIN CHEMICAL CLASSES OF GSR FROM VARIOUS AMMUNITION TYPES Zuzanna BRO EK-MUCHA, Grzegorz ZADORA, 2 Insttute of Forensc Research, Cracow, Poland 2 Faculty of Chemstry, Jagellonan

More information

POLYSA: A Polynomial Algorithm for Non-binary Constraint Satisfaction Problems with and

POLYSA: A Polynomial Algorithm for Non-binary Constraint Satisfaction Problems with and POLYSA: A Polynomal Algorthm for Non-bnary Constrant Satsfacton Problems wth and Mguel A. Saldo, Federco Barber Dpto. Sstemas Informátcos y Computacón Unversdad Poltécnca de Valenca, Camno de Vera s/n

More information

J. Parallel Distrib. Comput.

J. Parallel Distrib. Comput. J. Parallel Dstrb. Comput. 71 (2011) 62 76 Contents lsts avalable at ScenceDrect J. Parallel Dstrb. Comput. journal homepage: www.elsever.com/locate/jpdc Optmzng server placement n dstrbuted systems n

More information

A Probabilistic Theory of Coherence

A Probabilistic Theory of Coherence A Probablstc Theory of Coherence BRANDEN FITELSON. The Coherence Measure C Let E be a set of n propostons E,..., E n. We seek a probablstc measure C(E) of the degree of coherence of E. Intutvely, we want

More information

How To Plan A Network Wide Load Balancing Route For A Network Wde Network (Network)

How To Plan A Network Wide Load Balancing Route For A Network Wde Network (Network) Network-Wde Load Balancng Routng Wth Performance Guarantees Kartk Gopalan Tz-cker Chueh Yow-Jan Ln Florda State Unversty Stony Brook Unversty Telcorda Research kartk@cs.fsu.edu chueh@cs.sunysb.edu yjln@research.telcorda.com

More information

1. Math 210 Finite Mathematics

1. Math 210 Finite Mathematics 1. ath 210 Fnte athematcs Chapter 5.2 and 5.3 Annutes ortgages Amortzaton Professor Rchard Blecksmth Dept. of athematcal Scences Northern Illnos Unversty ath 210 Webste: http://math.nu.edu/courses/math210

More information

Fast degree elevation and knot insertion for B-spline curves

Fast degree elevation and knot insertion for B-spline curves Computer Aded Geometrc Desgn 22 (2005) 183 197 www.elsever.com/locate/cagd Fast degree elevaton and knot nserton for B-splne curves Q-Xng Huang a,sh-mnhu a,, Ralph R. Martn b a Department of Computer Scence

More information

FINANCIAL MATHEMATICS. A Practical Guide for Actuaries. and other Business Professionals

FINANCIAL MATHEMATICS. A Practical Guide for Actuaries. and other Business Professionals FINANCIAL MATHEMATICS A Practcal Gude for Actuares and other Busness Professonals Second Edton CHRIS RUCKMAN, FSA, MAAA JOE FRANCIS, FSA, MAAA, CFA Study Notes Prepared by Kevn Shand, FSA, FCIA Assstant

More information

) of the Cell class is created containing information about events associated with the cell. Events are added to the Cell instance

) of the Cell class is created containing information about events associated with the cell. Events are added to the Cell instance Calbraton Method Instances of the Cell class (one nstance for each FMS cell) contan ADC raw data and methods assocated wth each partcular FMS cell. The calbraton method ncludes event selecton (Class Cell

More information

Multi-Source Video Multicast in Peer-to-Peer Networks

Multi-Source Video Multicast in Peer-to-Peer Networks ult-source Vdeo ultcast n Peer-to-Peer Networks Francsco de Asís López-Fuentes*, Eckehard Stenbach Technsche Unverstät ünchen Insttute of Communcaton Networks, eda Technology Group 80333 ünchen, Germany

More information

The Load Balancing of Database Allocation in the Cloud

The Load Balancing of Database Allocation in the Cloud , March 3-5, 23, Hong Kong The Load Balancng of Database Allocaton n the Cloud Yu-lung Lo and Mn-Shan La Abstract Each database host n the cloud platform often has to servce more than one database applcaton

More information

Survey on Virtual Machine Placement Techniques in Cloud Computing Environment

Survey on Virtual Machine Placement Techniques in Cloud Computing Environment Survey on Vrtual Machne Placement Technques n Cloud Computng Envronment Rajeev Kumar Gupta and R. K. Paterya Department of Computer Scence & Engneerng, MANIT, Bhopal, Inda ABSTRACT In tradtonal data center

More information

Traffic-light a stress test for life insurance provisions

Traffic-light a stress test for life insurance provisions MEMORANDUM Date 006-09-7 Authors Bengt von Bahr, Göran Ronge Traffc-lght a stress test for lfe nsurance provsons Fnansnspetonen P.O. Box 6750 SE-113 85 Stocholm [Sveavägen 167] Tel +46 8 787 80 00 Fax

More information

Linear Circuits Analysis. Superposition, Thevenin /Norton Equivalent circuits

Linear Circuits Analysis. Superposition, Thevenin /Norton Equivalent circuits Lnear Crcuts Analyss. Superposton, Theenn /Norton Equalent crcuts So far we hae explored tmendependent (resste) elements that are also lnear. A tmendependent elements s one for whch we can plot an / cure.

More information

Traffic State Estimation in the Traffic Management Center of Berlin

Traffic State Estimation in the Traffic Management Center of Berlin Traffc State Estmaton n the Traffc Management Center of Berln Authors: Peter Vortsch, PTV AG, Stumpfstrasse, D-763 Karlsruhe, Germany phone ++49/72/965/35, emal peter.vortsch@ptv.de Peter Möhl, PTV AG,

More information

Sketching Sampled Data Streams

Sketching Sampled Data Streams Sketchng Sampled Data Streams Florn Rusu, Aln Dobra CISE Department Unversty of Florda Ganesvlle, FL, USA frusu@cse.ufl.edu adobra@cse.ufl.edu Abstract Samplng s used as a unversal method to reduce the

More information

Fuzzy Set Approach To Asymmetrical Load Balancing In Distribution Networks

Fuzzy Set Approach To Asymmetrical Load Balancing In Distribution Networks Fuzzy Set Approach To Asymmetrcal Load Balancng n Dstrbuton Networks Goran Majstrovc Energy nsttute Hrvoje Por Zagreb, Croata goran.majstrovc@ehp.hr Slavko Krajcar Faculty of electrcal engneerng and computng

More information

Time Value of Money. Types of Interest. Compounding and Discounting Single Sums. Page 1. Ch. 6 - The Time Value of Money. The Time Value of Money

Time Value of Money. Types of Interest. Compounding and Discounting Single Sums. Page 1. Ch. 6 - The Time Value of Money. The Time Value of Money Ch. 6 - The Tme Value of Money Tme Value of Money The Interest Rate Smple Interest Compound Interest Amortzng a Loan FIN21- Ahmed Y, Dasht TIME VALUE OF MONEY OR DISCOUNTED CASH FLOW ANALYSIS Very Important

More information

IDENTIFICATION AND CORRECTION OF A COMMON ERROR IN GENERAL ANNUITY CALCULATIONS

IDENTIFICATION AND CORRECTION OF A COMMON ERROR IN GENERAL ANNUITY CALCULATIONS IDENTIFICATION AND CORRECTION OF A COMMON ERROR IN GENERAL ANNUITY CALCULATIONS Chrs Deeley* Last revsed: September 22, 200 * Chrs Deeley s a Senor Lecturer n the School of Accountng, Charles Sturt Unversty,

More information

PAS: A Packet Accounting System to Limit the Effects of DoS & DDoS. Debish Fesehaye & Klara Naherstedt University of Illinois-Urbana Champaign

PAS: A Packet Accounting System to Limit the Effects of DoS & DDoS. Debish Fesehaye & Klara Naherstedt University of Illinois-Urbana Champaign PAS: A Packet Accountng System to Lmt the Effects of DoS & DDoS Debsh Fesehaye & Klara Naherstedt Unversty of Illnos-Urbana Champagn DoS and DDoS DDoS attacks are ncreasng threats to our dgtal world. Exstng

More information

Staff Paper. Farm Savings Accounts: Examining Income Variability, Eligibility, and Benefits. Brent Gloy, Eddy LaDue, and Charles Cuykendall

Staff Paper. Farm Savings Accounts: Examining Income Variability, Eligibility, and Benefits. Brent Gloy, Eddy LaDue, and Charles Cuykendall SP 2005-02 August 2005 Staff Paper Department of Appled Economcs and Management Cornell Unversty, Ithaca, New York 14853-7801 USA Farm Savngs Accounts: Examnng Income Varablty, Elgblty, and Benefts Brent

More information

Brigid Mullany, Ph.D University of North Carolina, Charlotte

Brigid Mullany, Ph.D University of North Carolina, Charlotte Evaluaton And Comparson Of The Dfferent Standards Used To Defne The Postonal Accuracy And Repeatablty Of Numercally Controlled Machnng Center Axes Brgd Mullany, Ph.D Unversty of North Carolna, Charlotte

More information

A Dynamic Energy-Efficiency Mechanism for Data Center Networks

A Dynamic Energy-Efficiency Mechanism for Data Center Networks A Dynamc Energy-Effcency Mechansm for Data Center Networks Sun Lang, Zhang Jnfang, Huang Daochao, Yang Dong, Qn Yajuan A Dynamc Energy-Effcency Mechansm for Data Center Networks 1 Sun Lang, 1 Zhang Jnfang,

More information

How Sets of Coherent Probabilities May Serve as Models for Degrees of Incoherence

How Sets of Coherent Probabilities May Serve as Models for Degrees of Incoherence 1 st Internatonal Symposum on Imprecse Probabltes and Ther Applcatons, Ghent, Belgum, 29 June 2 July 1999 How Sets of Coherent Probabltes May Serve as Models for Degrees of Incoherence Mar J. Schervsh

More information

Design and Development of a Security Evaluation Platform Based on International Standards

Design and Development of a Security Evaluation Platform Based on International Standards Internatonal Journal of Informatcs Socety, VOL.5, NO.2 (203) 7-80 7 Desgn and Development of a Securty Evaluaton Platform Based on Internatonal Standards Yuj Takahash and Yoshm Teshgawara Graduate School

More information

Feature selection for intrusion detection. Slobodan Petrović NISlab, Gjøvik University College

Feature selection for intrusion detection. Slobodan Petrović NISlab, Gjøvik University College Feature selecton for ntruson detecton Slobodan Petrovć NISlab, Gjøvk Unversty College Contents The feature selecton problem Intruson detecton Traffc features relevant for IDS The CFS measure The mrmr measure

More information

A new look at atomic broadcast in the asynchronous. crash-recovery model

A new look at atomic broadcast in the asynchronous. crash-recovery model A new look at atomc broadcast n the asynchronous crash-recovery model Sergo Mena André Schper École Polytechnque Fédérale de Lausanne (EPFL) Dstrbuted Systems Laboratory CH-1015 Lausanne, Swtzerland Tel.:

More information

EVALUATING THE PERCEIVED QUALITY OF INFRASTRUCTURE-LESS VOIP. Kun-chan Lan and Tsung-hsun Wu

EVALUATING THE PERCEIVED QUALITY OF INFRASTRUCTURE-LESS VOIP. Kun-chan Lan and Tsung-hsun Wu EVALUATING THE PERCEIVED QUALITY OF INFRASTRUCTURE-LESS VOIP Kun-chan Lan and Tsung-hsun Wu Natonal Cheng Kung Unversty klan@cse.ncku.edu.tw, ryan@cse.ncku.edu.tw ABSTRACT Voce over IP (VoIP) s one of

More information

Application of Multi-Agents for Fault Detection and Reconfiguration of Power Distribution Systems

Application of Multi-Agents for Fault Detection and Reconfiguration of Power Distribution Systems 1 Applcaton of Mult-Agents for Fault Detecton and Reconfguraton of Power Dstrbuton Systems K. Nareshkumar, Member, IEEE, M. A. Choudhry, Senor Member, IEEE, J. La, A. Felach, Senor Member, IEEE Abstract--The

More information

LIFETIME INCOME OPTIONS

LIFETIME INCOME OPTIONS LIFETIME INCOME OPTIONS May 2011 by: Marca S. Wagner, Esq. The Wagner Law Group A Professonal Corporaton 99 Summer Street, 13 th Floor Boston, MA 02110 Tel: (617) 357-5200 Fax: (617) 357-5250 www.ersa-lawyers.com

More information

Performance Analysis of Energy Consumption of Smartphone Running Mobile Hotspot Application

Performance Analysis of Energy Consumption of Smartphone Running Mobile Hotspot Application Internatonal Journal of mart Grd and lean Energy Performance Analyss of Energy onsumpton of martphone Runnng Moble Hotspot Applcaton Yun on hung a chool of Electronc Engneerng, oongsl Unversty, 511 angdo-dong,

More information