Performance of a Genetic Algorithm for Solving the Multi-Objective, Multimodal Transportation Network Design Problem

Size: px
Start display at page:

Download "Performance of a Genetic Algorithm for Solving the Multi-Objective, Multimodal Transportation Network Design Problem"

Transcription

1 , pp Performance of a Genetc Algorthm for Solng the Mult-Obecte, Multmodal Transportaton Netork Desgn Problem Tes Brands and Erc C. an Berkum Centre for Transport Studes, Unersty of Tente Postbus 27, 7500AE Enschede, the Netherlands t.brands@utente.nl, e.c.anberkum@utente.nl Abstract The optmzaton of nfrastructure plannng n a multmodal netork s defned as a multobecte netork desgn problem, th accessblty, use of urban space by parkng, operatng defct and clmate mpact as obectes. Decson arables are the locaton of park and rde facltes, tran statons and the frequency of publc transport lnes. For a case study the Pareto set s estmated by the Non-domnated Sortng Genetc Algorthm (NSGA-II). Such a Pareto set s one specfc outcome of the optmzaton process, for a specfc alue of the parameters generaton sze, number of generatons and mutaton rate and for a specfc outcome of the Monte Carlo smulaton thn NSGA-II. Smlar ssues exst for many other metaheurstcs. Hoeer, hen appled n practce, a polcy maker desres a result that s robust for these unknon aspects of the method. In ths paper Pareto sets from arous runs of the NSGA-II algorthm are analyzed and compared. To compare the alues of the decson arables n the Pareto sets, ne methods are necessary, so these are defned and appled. The results sho that the dfferences concernng decson arables are consderably larger than the dfferences concernng obectes. Ths ndcates that the randomness of the algorthm may be a problem hen determnng the decsons to be made. Furthermore, t s concluded that aratons caused by dfferent parameters are comparable th the aratons caused by randomness thn the algorthm. Keyords: Mult-obecte optmzaton, genetc algorthm, NSGA-II, multmodal passenger transportaton netorks, Pareto set comparson. Introducton Hghly urbanzed regons n the orld noadays face ell-knon sustanablty problems n traffc systems, lke congeston, use of scarce space n ctes by ehcles and the emsson of greenhouse gases. A shft from car to publc transport (PT) modes s lkely to alleate these problems. Extenson of the PT netork may contrbute to ths shft by enhancng the qualty of PT trps. Hoeer, nestments n PT nfrastructure requre large fnancal resources. Better utlzaton of the exstng PT nfrastructure thout the need for bg nestments n large nfrastructural deelopments may be acheed by facltatng an easy transfer from prate modes (bcycle and car) to PT modes (bus, tram, metro, tran). Ths transfer can be made easer by netork deelopments that enable multmodal trps, lke openng ne Park and Rde (P&R) facltes, openng ne tran statons or changng exstng or openng ne PT serce lnes. When such transportaton netork deelopments are planned, the decson s often based on an ealuaton of a fe pre-defned scenaros. The composton of the scenaros s usually based on expert udgment, and the assessment s usually based on a mult crtera analyss. ISSN: IJT Copyrght c 204 SERSC

2 Hoeer, there s a far chance that the scenaro that s selected as the best can stll be mproed. For ths reason applyng an optmzaton technque s preferred. In ths paper, a multobecte netork optmzaton approach s adopted, hch enables us to dentfy trade-offs beteen obectes (see also []), because the feld of transportaton plannng s, lke many other felds, a typcal example of a stuaton that noles more than one obecte [2]. Netork desgn problems (NDPs) hae receed a lot of attenton n the lterature, n many dfferent ersons. One subclass of problems s the transt netork desgn problem, hch has been studed n arous ays, as reeed by [3]. Ths ncludes greedy algorthms, eolutonary algorthms and desgn meetngs nolng expert udgments. Another subclass of problems s the unmodal road netork desgn problem, hch has also been studed n arous ays [4]. Applcatons n a multmodal context are less common, but they do exsts, for example road lnk capacty and bus routes [5] or prcng of prate and publc lnks [6]. The NDP s consdered a mult-obecte problem n only a lmted number of cases. Examples of obectes beng consdered smultaneously are trael tme and constructon costs [7], trael tme and trael tme araton [8], trael tme and CO emssons [9, 0] and socal elfare mproement, reenue generaton and equty []. Fnally, [5] smultaneously maxmze user benefts, passenger share of the bus mode, serce coerage, hle mnmzng aerage generalzed trael cost for a bus mode trp. To our knoledge [5] s the only paper combnng mult-obecte optmzaton th a multmodal NDP, consderng both ne street constructon and lane addtons / allocatons as ell as redesgn of bus routes. The focus s on the deelopment of the metaheurstcs to sole the problem: the performance of a hybrd genetc algorthm and a hybrd clonal selecton algorthm are compared, usng multple test netorks. In the loer leel, car and bus are dstngushed as separate modes. A number of papers appled metaheurstcs for solng the transt netork desgn problem, for example [2-4] use a genetc algorthm, [5] uses smulated annealng, [6] tests both scatter search as a genetc algorthm and [7] uses a metaheurstc called GRASP (Greedy Randomzed Adapte Search Procedure). The randomness n such algorthms s usually addressed by dong multple runs and presentng aerage alues, for example hen comparng arous (arants of) soluton algorthms (as s done for example n [2]). An aspect that s less addressed s the araton of the results thn one arant of a soluton algorthm, th respect to the araton of the results beteen to arants of a soluton algorthm. Another aspect s the comparson of decson arables of optmzaton outcome. To be able to compare arous results n the case of mult-obecte optmzaton, one must be able to compare Pareto sets th each other, both concernng obecte alues as ell as alues of decson arables. Frstly, the contrbuton of ths paper s to formalze methods to compare Pareto sets based on the alues of decson arables. Secondly, the mpact of randomness and parameter settngs n the applcaton of a metaheurstc lke NSGA-II on the outcome of the optmzaton process s assessed, for the case of a netork desgn problem th multmodal decson arables. Therefore, exstng ndcators to compare the obecte alues of dfferent Pareto sets are used, n addton to the nely deeloped methods to compare the alues of decson arables. Secton 2 defnes the problem n more detal and descrbes the soluton method appled. Secton 3 defnes methods to compare Pareto sets th each other, to be used n Secton 4, here the outcomes of a real orld case study are analyzed. The case study s descrbed and the alues of arous outcomes of the optmzaton process are compared th each other. Ths leads to conclusons n the fnal secton. 2 Copyrght c 204 SERSC

3 2. Problem Defnton 2.. B-leel Problem The transportaton netork desgn problem s often soled as a b-leel optmzaton problem, to correctly ncorporate the reacton of the transportaton system users to netork changes, as s argued by [8-20]. In our research, the netork desgn problem s regarded as a b-leel system as ell (see Fgure ). The upper leel represents the behaor of the netork authorty, optmzng system obectes. In the loer leel the traelers mnmze ther on generalzed costs (e.g., trael tme, cost), by makng nddually optmal choces n the multmodal netork, consderng arety n trael preferences among traelers. The netork desgn n the upper leel nteracts th the behaor of the traelers n the netork: the loer leel. Ths s put nto operaton by a transport model, hch assumes a stochastc user equlbrum (no drer can unlaterally change routes to mproe hs/her perceed generalzed trael costs). For any netork desgn the planner chooses, the transport model yelds a netork state (e.g., trael tmes and loads), from hch the alues of all obecte functons can be dered. The equlbrum n the loer leel s a constrant for the upper leel problem. Upper leel: Mnmzaton system obectes concernng sustanablty Netork state Loads, speeds Loer leel: User equlbrum problem: multmodal assgnment Netork desgn: Multmodal facltes 2.2. Netork and Demand Defnton Fgure. The B-leel Optmzaton Problem The multmodal transportaton netork s defned as a drected graph G, consstng of node set N, lnk set A, a lne set L and a stop set U (see Table ). For each lnk a one or more modes are defned that can traerse that lnk th a certan speed and capacty: the lnk characterstcs C. Orgns r R and destnatons s S a are a subset of N and act as orgns Copyrght c 204 SERSC 3

4 Table. Notaton NSGA-II MOOP NDP PT P&R IC Table 2. Abbreatons Non-domnated Sortng Genetc Algorthm (second erson) Mult-Obecte Optmzaton Problem Netork Desgn Problem Publc Transport Park and Rde Intercty (tran) 4 Copyrght c 204 SERSC

5 and destnatons. Total fxed transportaton demand q s stored n a matrx th sze R S. rs Furthermore, transt serce lnes l L are defned as ordered subsets A thn A and can be l stop serces or express serces. PT flos can only traerse transt serce lnes. Transt statons or stops u U are defned as a subset thn N. Consequently, a lne l traerses seeral stops. The trael tme beteen to stops and the frequency of a transt serce lne l are stored as lne characterstcs C. Access / egress modes and PT are only connected l through these stops. A combnaton of usng a specfc access mode, PT and a specfc egress mode s defned as a mode chan M. Whether a lne calls at a stop u or not, s ndcated by stop characterstcs C. All together, the transportaton netork s defned by G ( N, A, L, U ), u here A, L and U are further specfed by C, C and C. a l u 2.3. Optmzaton Problem We defne a decson ector y (or a soluton), that conssts of V decson arables: y y,, y,, y V. Y s the set of feasble alues for the decson ector y (also called decson space). The obecte ector Z (consstng of W obecte functons, Z Z,, Z,, Z W ) depends on the alue of the decson ector y. Eery Z s part of the so called obecte space, and n prncple Z may be any alue n W, but dependng on ts meanng, an obecte functon may be subect to natural bounds. The resultng mult-obecte optmzaton problem (MOOP) s defned n program -5. The loer leel program n 3-5 s based on the formulaton n [2]. The characterstcs C, C and l C of A, L and U depend on the decson ector y. These characterstcs defne the u multmodal transportaton netork G. A fxed demand rs q s assgned to ths transportaton netork, such that satsfes the constrants (2-3). Feasblty constrant (2) should be satsfed to defne a transportaton netork that s physcally thn reach. Constrant (3) represents the loer leel optmzaton problem, that optmzes modal splt and flos n the netork from the traeler s perspecte and has equatons (4-5) as constrants. Constrant (4) ensures that the sum of car route flos and transt demand equal total demand. Constrant (5) s a non-negatty constrant for car route flos. Note that PT route flos are nonnegate by the defntons n constrants (6-7). (6) ensures poste PT demand, but bounded by the total demand. (7) dstrbutes PT demand oer arous mode chans and PT routes, usng the fractons n constrants (2-3): (3) defnes the route fractons for PT trps per mode chan: the PT model contans multple routng, here trps are dstrbuted among sensble routes usng a logt model. Constrant (2) defnes the mode chan fractons for PT trps. Constrant (0) calculates the OD based generalzed costs per mode chan as a eghted aerage of route costs, to correctly take nto account combned atng costs of multple transt lnes. Constrant (9) calculates the logsum for the mode chans, to represent OD based PT costs. Constrants (8) and () defne the relaton beteen lnk costs and route costs, for car and for PT. Fnally, constrants (4-5) defne the ehcle and passenger lnk flos based on car and PT route flos. Note that PT routes may contan car lnks, n the case of the use of a P&R faclty n the route (see constrant 4). On the other hand, car routes neer nclude PT passenger lnks, as can be obsered n constrant (8). a Copyrght c 204 SERSC 5

6 Based on the resultng ehcle and passenger flos n the netork G, an obecte ector s calculated for a decson ector y n a general formulaton (). The case specfc formulaton of obecte functons s gen separately n Secton Mult-obecte Optmzaton: Pareto Optmalty Mathematcally, the concept of Pareto optmalty s as follos. If e assume to decson ectors y, y Y, then y s sad to strongly domnate y ff Z ( y ) Z ( y ) ' ' (also rtten as y y ) Addtonally, ' y s sad to coer or eakly domnate y ff Z ( y ) Z ( y ) (also rtten as y y ). All solutons that are not ' ' ' eakly domnated by another knon soluton are possbly optmal for the decson maker: these solutons form the Pareto-optmal set P. ' 6 Copyrght c 204 SERSC

7 2.5. Soluton Method The MOOP defned n (-5) s an NP-hard problem, snce the b-leel lnear programmng problem s already an NP-hard problem [22]. Therefore, the problem s computatonally too expense to be soled exactly for larger netorks, so e rely on a heurstc. For solng mult-obecte problems, the class of genetc algorthms s often used, because these algorthms hae a lo rsk of endng up n a local mnmum, do not requre the calculaton of a gradent and are able to produce a derse Pareto set [23]. More specfcally, e use the NSGA-II algorthm, deeloped n [24]. NSGA-II has been successfully appled by researchers to sole mult-obecte optmzaton problems n traffc engneerng and proed to be effcent for ths type of problems [2, 8]. It optmzes multple obectes smultaneously, searchng for a set of non-domnated solutons,.e. the Pareto optmal set. It s a genetc algorthm, based on the prncples of natural selecton thn eoluton, combnng solutons to ne solutons (crossoer), here the solutons th a hgher ftness alue hae a larger chance to sure oer orse solutons. In the next generaton, these enhanced solutons are recombned agan, untl no progress s made any more or untl the maxmum number of generatons s reached. Wthn NSGA-II, the matng selecton s done by bnary tournament selecton th replacement. In addton to ths matng process, a random mutaton operator s appled to a (small) fracton of solutons from each generaton, to promote the exploraton of unexplored regons n the decson space. These aspects of the algorthm make that the result s an approxmaton of the true Pareto set. The random processes n the algorthm cause that from sub sequental runs of the same algorthm, dfferent Pareto sets result. Furthermore, NSGA-II contans eltsm, to presere good solutons n an arche. If the number of non-domnated solutons gros bgger than the arche, the arche only contans the best non-domnated solutons based on the defned ftness alue. The ftness alue s calculated n to steps. In the frst step (non-domnated sortng), the solutons are ranked based on Pareto domnance. All solutons n the Pareto set recee rank. In the next step, these solutons are extracted from the set and all Pareto solutons n the remanng set recee rank 2, etc. In the second step, the solutons are sorted thn these ranks based on ther crodng dstance. Crodng dstance calculaton requres sortng of the populaton accordng to each obecte alue. The extreme alues for each obecte are assgned an nfnte alue, assurng that these alues sure. All ntermedate solutons are assgned a alue equal to the dfference n the normalzed functon alues of to adacent solutons. Concludng, the crodng dstance alue (and thus the ftness alue) s hgher f a solutons s more solated, promotng a more derse Pareto optmal set. The algorthm n steps (for more detals, the reader s referred to [24]): Step : Intalzaton: Set populaton sze p, hch s equal to the arche sze a, the maxmum number of generatons H, and generate an ntal populaton. Set. 0 h 0 and. 0 Step 2: Ftness assgnment: Combne arche and offsprng, formng and h h h h h calculate ftness alues of solutons by domnance rankng and croded dstance sortng. Step 3: Enronmental selecton: Determne ne arche h by selectng the best a solutons out of based on ther ftness. h Copyrght c 204 SERSC 7

8 Step 4: Termnaton: If h H or another stoppng crtera s satsfed, determne Pareto set P from the set of all calculated solutons 0 H (non-domnated solutons). Step 5: Matng selecton: Perform bnary tournament selecton th replacement on h to determne matng pool of parents. h Step 6: Varaton: Apply recombnaton and mutaton operators to the matng pool g to create offsprng. h Set h h and go to step Output Defntons The set s defned as all decson ectors (or solutons) that are calculated durng one optmzaton process, so,,, 2 N P y y y H a. The set of N solutons s defned as the Pareto set resultng from process, hch ncludes all non-domnated solutons th respect to all solutons n s no y P : y y, y. ' ', or mathematcally there P s the -th outcome of our MOOP (): one approxmaton of the Pareto-optmal set. The -th executon of the algorthm s also denoted as run. Pareto set P of all calculated solutons { J } s denoted as the superset, and s the best knon Pareto set based on J runs. Please recall that eery element or soluton y represents one transportaton netork desgn that s defned by the alues of V decson arables n,,,,, 2 V. For eery element y y y y y y : y a ector of obecte functons Z ( y ) s calculated, that conssts of W obecte functons, denoted by ndex : Z ( y ) Z ( y ), Z ( y ),, Z ( y ) W. 3. Comparng Multple Optmzaton Outcomes When comparng Pareto sets, t s common to assess the qualty of the Pareto sets by ndcators. These ndcators are usually based on the obecte alues of the solutons n the Pareto set, to measure ether the extent to hch the set s close to the real Pareto-optmal front or the extent to hch the set s derse [23]. Hoeer, another aspect that s releant hen comparng dfferent outcomes of a randomzed algorthm lke NSGA-II (and many other MO optmzaton algorthms), s the extent to hch to Pareto sets are smlar n the decson space. If a Pareto set s used as a decson support tool, the leel of uncertanty n the results s mportant for the decson maker. Especally the uncertanty n the decson space s of mportance, because that s the range he or she can choose from,.e., ths determnes f he or she has proper nformaton to make the rght decson. The parameter alues n the algorthm (n the case of NSGA-II, H a and ) may be another source of uncertanty, hch s ndcated by the senstty of the results for these parameters. To be able to assess ths smlarty formally for our case study n Secton 4, n the next secton comparson ndcators are defned. 8 Copyrght c 204 SERSC

9 3.. Methods for Comparson A dstncton may be made based on to man characterstcs of the method. The frst s a dstncton beteen the number of Pareto sets to be compared: a unary ndcator can be calculated for one Pareto set on ts on, a bnary ndcator can be calculated for to Pareto sets [25]. The so called attanment functon can be calculated for J Pareto sets. The second dstncton s hether the method compares the Pareto sets n the obecte space or n the decson space: the frst type of method s calculated based on the obecte functon alues Z of the solutons n the set(s), hle the second type of method s calculated based on the decson alues y of the solutons n the set(s). For comparson n the obecte space, a selecton of ndcators from the lterature s presented. For comparson n the decson space, ne ndcators are defned. An oere of the ndcators used n ths paper s gen n Table 3. These ndcators use a dstance functon, hch can be any dstance functon defned for ' to solutons as ( d y, y ). In ths paper the dstance functon s defned as n formula 6. ' V ' ' ' '. (6) d ( y, y ) y y 4. Case Study We apply the mult-obecte optmzaton frameork to a real lfe multmodal transportaton netork, stuated n the area around the Dutch ctes of Amsterdam and Haarlem (see Fgure 2). Frst, the obectes and decson arables for ths specfc case are descrbed. After that, the dfferent runs of the NSGA-II algorthm are lsted, ncludng ther scores on unary performance ndcators. Fnally, the runs are compared th each other by usng the bnary ndcators and the attanment functon. Fgure 2. Map of the Study Area, Contanng Transportaton Zones, Ralays and Roads Copyrght c 204 SERSC 9

10 Table 3. Oere of Indcators to Compare Pareto Sets Formula Type Descrpton N Unary Cardnalty of Pareto set. The bgger a Pareto set, the more complex t s to analyze for a decson maker. Furthermore, th more solutons n the set t s easer to fulfll other crtera, lke hyperolume. M IN ( P ) m n Z ( y ), y P M A X ( P ) m a x Z ( y ), R M I k S S C ( P ) z ( y) 2 ( P ) S S C y P M IN ( P ) M IN ( P ) P k M A X ( P ) M IN ( P ) z ( y) Fg. 3: hyperolume 2-D sualzaton ' C T S ( P, P ) ' ' y P ; y P : y y ' ' ' A D ( P, P ) d ( y, y ) ' N N N N ' ' ' ' A N D ( P, P ) m n (, ) d y y ' ' N '.. N N N V ' ' A F D ( P, P ) V N N ( y ), th 0 f y 0 ( y ) o th er se ', th ' Unary, obecte space Unary, obecte space Bnary, obecte space Bnary, decson space Bnary, decson space Bnary, decson space The mnmum obecte alue attaned by Pareto set for eery obecte functon, th respect to the range Z coered by the superset. For that, the mnmum and maxmum alues are determned for the superset, as ell as the mnmum alue for set. Ths ndcates the extent to hch set coers the entre Pareto front for eery obecte dmenson. The space coerage. Implemented as n [26], also knon as S-metrc or hyperolume. In the 2-dmensonal case t determnes the area that s coered by the Pareto set th respect to a reference pont (the star n fg. 3). The reference pont represents the upper bound of all obectes: the reference pont s defned such that t s domnated by all solutons n the Pareto set. Because the true maxmum alues of the obecte functons are not knon, e choose a conserate pont, based on the ealuated solutons. In the 3 dmensonal case area s replaced by olume, and n the more dmensonal case by hyperolume. The set coerage or C-metrc, see [27]. The leel n ' hch the solutons n P are eakly domnated by at least one soluton n set P, so a hgher alue ndcates ' a better set coerage of P oer P. The aerage dstance (n the decson space) beteen the ' solutons n to Pareto sets P and P (here may also be equal to ). Ths ndcator s used to check hether the araton thn a Pareto set s dfferent from the araton beteen Pareto sets. The dstance of an element n P to the closest element (.e. ' most smlar n decson space) n Pareto set P, aeraged oer all elements P. A lo alue ndcates hgher smlarty (the extreme case A N D ( P, P ) 0 holds). Comparson of fractons of nonzero decson arables. The fracton of solutons n Pareto set that hae a poste alue for decson arable y characterzes the decson space of the set: decson arable y represents the exstence of a measure n the transportaton netork, so a hgher fracton for arable y mples a larger ablty to attan Pareto solutons hen that arable has a poste alue. The functon defnes the nonzero relaton. To ndcate the dfference beteen Pareto set and, the 0 Copyrght c 204 SERSC

11 J ( Z ) I ( y P : y Z ) J dfferences beteen these fractons are calculated, aeraged oer all V decson arables. Compare Attanment functon, as defned n [25]. For each obecte J Pareto ector Z ths functon returns the probablty that Z s sets, attaned by a fracton of J Pareto sets. It s approxmated obecte on the bass of the approxmaton set sample. Functon space I returns f ts argument s true and 0 f t s false. 4.. Obectes In ths paper e consder 4 polcy obectes related to sustanablty, concernng accessblty, use of urban space by parkng, clmate mpact and costs. These obectes and ther operatonalzaton are shon n table 4. Varables that are not defned earler, are defned belo the table. All 4 obectes are to be mnmzed. In our case of a fxed total demand, total trael tme ges the same results as aerage trael tme ould hae gen. The urban space used by parkng s represented by the number of car trps to or from zones that are classfed as hghly urban, because such a trp requres a parkng space that cannot be used for other urban land uses. These alternate land uses ge addtonal alue to property [28]. CO 2 emssons are calculated usng the ARTEMIS traffc stuaton based emsson model, here emsson factors for cars on a lnk depend on free flo speed and olume / capacty rato [29]. Emsson factors for PT depend on ehcle type. Operatng defct of the PT system s formulated as an obecte, rather than as a budget constrant, to prode explct nsght n the relaton beteen costs and other obectes. Cost parameters follo from Dutch PT operatng practce Decson Varables Decson arables n ths case study are related to transfer facltes or to PT facltes. For eery potental netork deelopment, a decson arable y s defned n adance (see Table 5). Netork deelopments are only ncluded as a canddate locaton f spatal and capacty constrants are met. For example, a P&R locaton s only opened f the correspondng staton s sered by PT. The characterstcs of lnks, lnes and stops that are not canddate locatons are fxed at one alue. Furthermore, the car and bcycle netorks are assumed to be fxed. In ths case, the feasble regon Y contans approxmately 0 9 possble decson ectors. Frequency of PT Serce Lnes A part of the decson ector y represents frequences of PT serce lne, ncludng tran lnes (local trans stop at all statons they pass through, Intercty (IC) trans or express trans only stop at IC statons) and man bus lnes (th multple stops). In our case, the frequences can be chosen from 0, 2, 4, 6, so the alue may also be 0: the lne does not exst n that case. To achee comparable alues of y oer the entre the decson space, the correspondng alues that y are normalzed as 0,,,. Tram lne extensons are defned as a bnary arable, ndcatng hether the lne s extended or not (th the orgnal frequency). A feeder bus netork s aalable n the netork, but s assumed to be fxed n ths case. Exstence of Tran Statons For the exstence of tran statons y s defned as a bnary arable, that ndcates hether transt ehcles call at a stop u or not (.e. exchange beteen the alk and bcycle netorks Copyrght c 204 SERSC

12 and the PT netork). Note that for bus stops and for exstng tran statons the alue of fxed to : these PT stops can n all solutons be reached by bcycle and alkng. Express status of tran staton u For the express tran status of tran statons, y s defned as a bnary arable, that ndcates hether transt ehcles of express lnes call at a stop u or not. Note that at exstng express stops ths arable s fxed to. Table 4. Obectes and ther Operatonalzaton n Words and n Mathematcal Formulaton y s Park and Rde faclty at stop u For Park and Rde, y s defned as a bnary arable that ndcates hether t s possble to reach a stop by car, by changng the characterstcs of the lnk connectng the stop to the access netork. Note that there are also exstng P&R facltes that are fxed to,.e., PT stops that can n all solutons be reached by car Pareto Set Approxmatons The NSGA-II algorthm as executed 3 tmes n total (see Table 6). The frst 5 runs ere executed th dentcal parameter settngs, to nestgate the mpact of randomness n the algorthm. After that, n to more groups of 3 runs, 2 dfferent combnatons of parameter settngs ere chosen. Fnally, to unque runs ere executed, th a dfferent mutaton rate. A set of runs th the same parameter settngs s further denoted as a group of runs. The groups are hghlghted n Table 6. All parameter settngs ere chosen n such a ay that 2 Copyrght c 204 SERSC

13 computaton tmes ere stll reasonable (one eek per run), hle conergence as reached or almost reached n terms of attaned hyperolume. Furthermore, for a far comparson, the total computaton tme (ndcated by the total number of calculated solutons H ) as a approxmately the same for all runs. Decson arable ndex Table 5. The Defnton of 26 Decson Varables n the Case Study Possble alues of y Represents real alue Descrpton of decson arable 2, 2, 4 Frequency of ntercty tran Amsterdam - Haarlem Leden 2 0, 0, 2 Frequency of local tran Amsterdam - Haarlem Leden 3 2 0,,, 0, 2, 4, Frequency of ne tran lne Utgeest - Haarlem - Amsterdam Sloterdk - Zud Blmer 4 0, 0, 2 Frequency of tran lne Utgeest - Haarlem - Amsterdam Centraal Fracton of nonzero alues n superset 00% 5 0, 0, 2 Frequency of tran lne Haarlem Leden 4% 6 2 0,,, 0, 2, 4, ,,, 0, 2, 4, ,,, 0, 2, 4, ,,, 0, 2, 4, ,,, 0, 2, 4, ,,, 0, 2, 4, Frequency of bus lne 76: Amsterdam Zud Haarlem Frequency of bus lne 83: IJmuden A dam Sloterdk Frequency of ne bus lne Hoofddorp - Amsterdam West - Sloterdk Frequency of bus lne 75: Haarlem - Amsteleen - Amsterdam Zudoost Frequency of bus lne 80: Zandoort - Haarlem - Amsterdam Marnxstraat Frequency of ne bus lne 'Westtangent': Amsterdam Sloterdk Amsterdam West Schphol 2 0, Exstence Extenton of tram 3 n Geuzeneld 42% 3 0, Exstence Extenton of express tram Amsteleen to Uthoorn 62% 4 0, Exstence Ne tran staton Halfeg-Zanenburg % 5 0, Exstence Ne tran staton Haarlem Zud 36% 6 0, Exstence Ne tran staton Amsterdam Geuzeneld 0% 7 0, no ICs, ICs Intercty status of staton Hoofddorp 88% 8 0, no ICs, ICs Intercty status of staton Heemstede-Aardenhout 7% 9 0, no ICs, ICs Intercty status of staton Amsterdam Lelylaan 22% 20 0, no ICs, ICs Intercty status of staton Duendrecht 45% 2 0, Exstence P&R at tran staton Halfeg-Zanenburg 0% 22 0, Exstence P&R Velsen Zud (at bus lne 83) 30% 23 0, Exstence P&R at tram staton Amsteleen Oranebaan 2% 24 0, Exstence P&R at tran staton Geuzeneld and tram 3 5% 25 0, Exstence P&R Badhoeedorp at ne bus lne 2% 26 0, Exstence P&R Schphol Noord at bus lnes 'Westtangent' and 'Zudtangent' 6% 75% 5% 76% 99% 53% 6% 74% 62% 30% Copyrght c 204 SERSC 3

14 Table 6. Parameter Settngs and Unary Indcators of the 3 Runs of the Optmzaton Process Set number arche sze a Number of generatons H Mutaton rate Number of solutons n set N Hyperolume S S C ( P ) (*0 2 ) Unary Performance Indcators All 3 runs ere able to attan consderably larger hyperolumes (Table 6) than the random startng populatons (the aerage alue of hyperolume coered by the startng J populatons to as 7.3*0 2 ). Furthermore, the nddual runs are not too far 0 0 from the best knon Paretoset, the superset P, takng nto account that the superset contans a larger number of solutons than the nddual runs. When e compare the number of attaned Pareto solutons, only run 2 shos a large dfference th all other runs. The araton thn the number of Pareto solutons s smlar thn groups of runs and beteen groups of runs. The same holds for the hyperolume that s coered by the Pareto sets. In addton, run has a relately lo alue for hyperolume, hch seems to be an outler resultng from the random optmzaton process. Next, the mnmum alues attaned by the dfferent runs R M I ( P ) for all 4 obecte functons are compared. Some obseratons from Fgure 4 are: - Trael tme has the most (relate) araton from all obectes - Run 2 does not prode (near) optmal solutons for any obecte - Run 7 performs excellent, acheng 3 mnmal alues and aerage alue Although no hard conclusons can be dran based on one optmzaton outcome, the unary 2 ndcators sho that P has a orse performance than the other Pareto sets. Ths s also 2 shon by the set coerage th respect to P : the aerage alue oer, 2 of 2 C T S ( P, P ) s 0.94, hle the aerage alue oer, 2 of 2 C T S ( P, P ) s Ths ndcated that a mutaton rate of 0. (as as appled n run 2) s too hgh for ths type of problem. 4 Copyrght c 204 SERSC

15 RMI(P) Internatonal Journal of Transportaton 4% 2% 0% 08% 06% 04% Trael tme Urban space used Operaton defct CO2 emssons 02% 00% Fgure 4. The Relate Values R M I ( P ), for 3 Runs and for 4 Obectes 4.5. Comparng Pars of Pareto Sets In ths secton a parse comparson beteen Pareto sets s made. Frstly, Pareto sets 2 and 3 are plotted as an example n fg. 5. Ths shos that the sets hae comparable shapes, but the extremes are dfferent, and one Pareto set contans more solutons n certan areas than the other set. Next, the ndcators defned n Secton 3. are used to compare the Pareto sets. ' Table 7 shos the alues of ( A D P, P ) for each par of Pareto sets. The dagonal of the table contans the alues for the aerage dstance thn the same Pareto set. No clear dfferences can be obsered n the table. Hoeer, there s a small dfference beteen cases here ' and cases here ' : oerall, these aerage alues are 6.5 resp Furthermore, there s a small dfference beteen the aerage alue thn groups of runs (6.7) and beteen groups of runs (6.9). These dfferences seem too small to be substantal, although one ould expect that a par of Pareto sets th the same parameter settngs ould hae more smlartes than a par of Pareto sets th dfferent parameter settngs. Fgure 5. 2 Dmensonal Plots of To Dfferent Outcomes of the Optmzaton Process, for Obecte and 3 (Left) and for Obecte 2 and 4 (Rght) Copyrght c 204 SERSC 5

16 Table 7. The Aerage Dstance beteen Pareto Solutons ' A D ( P, P ) Table 8 shos the alues of ' ( A N D P, P ) for each par of Pareto sets. Note that the ' dagonal only contans zeros: n the same Pareto set an dentcal soluton th ( d y, y ) = 0 can alays be found. Another obseraton s that the ndcator s not symmetrcal,.e., ' ' (, ) (, A N D P P A N D P P ) : a soluton y n set that acts as nearest soluton for soluton ' y may not hae soluton ' ' y as nearest soluton n set. Agan, only mnor dfferences can ' ' be obsered n the table. Hoeer, the absolute alues of ( A N D P, P ) hae an mplcaton: on aerage, the closest soluton n another Pareto optmal set can be found at a dstance of 2.3. Ths mples that on aerage more than 2 of 26 the decson arables are dfferent f a soluton from another random outcome of the optmzaton process s used, hch s a substantal dfference. Furthermore, also n ths ndcator there s a small dfference beteen the aerage alue thn groups of runs (2.2) and beteen groups of runs (2.3). These dfferences seem too small to substantal. ' Table 8. The Aerage Dstance to the Nearest Pareto Soluton ' A N D ( P, P ) Copyrght c 204 SERSC

17 Table 9. The Dfference beteen Fractons of Nonzero Decson Varables n the ' Pareto Sets A F D ( P, P ) % 2.5% 8.7% 2.0% 9.3% 3.7%.7% 4.6% 9.4%.7% 9.8% 9.8% 2 8.3% 5.6% 7.7% 7.5% 8.3% 9.9% 9.7% 7.8% 5.9% 9.5% 6.7% 3 7.0% 6.4% 8.9% 8.7% 9.5% 9.4% 5.4% 6.%.7% 6.9% 4 6.8% 8.3% 9.6% 0.5% 0.0% 5.9% 6.5% 0.8% 6.2% 5 7.5% 7.6% 8.% 7.4% 7.6% 4.5%.7% 8.5% 6 8.0% 7.6% 9.7% 7.% 7.8% 9.% 7.4% 7 9.6% 9.0% 9.3% 6.7% 0.8% 9.2% 8 7.% 0.0% 8.4% 9.7% 8.3% 9 0.2% 7.%.3% 9.9% 0 6.8% 7.7% 5.8% 0.5% 7.% 2 9.4% ' Table 9 contans the alues for ( A F D P, P ). Relately large dfferences beteen pars of Pareto sets can be obsered. The absolute alue of ths ndcator has an mplcaton as ell: on aerage, the fracton of Pareto solutons n hch a certan decson arable s Pareto optmal, dffers 8.7%. Ths can be consdered as substantal, takng nto account that some decson arables are (almost) alays equal to 0, or (almost) alays nonzero (see table 5). Furthermore, also n ths ndcator there s a small dfference beteen the aerage alue thn groups of runs (8.4%) and beteen groups of runs (8.8%). These dfferences seem too small to be substantal Comparng Multple Pareto Sets Fnally, Fgure 6 shos the attanment plots of the 5 dentcal parameter settngs, for the obecte combnatons of total trael tme and CO 2 emssons resp. urban space used and operatng beneft. For the frst combnaton of obectes, some clear araton n results can be obsered, manly n the extreme ends of the Pareto set. Ths could already be seen n Fgure 5. The second combnaton of obectes only shos a small boundary of uncertanty, also because the tradeoff regon of the Pareto set s much bgger th these obectes. 5. Conclusons In ths paper the NSGA-II algorthm s appled to the multmodal transportaton netork desgn problem, as an example applcaton of a mult-obecte eolutonary algorthm. The performance of seeral runs of the optmzaton process s llustrated by a case study. The dfferences beteen multple random outcomes of the algorthm are analyzed, both th a sngle parameter settng and th dfferent parameter settngs. For ths analyss, arous ndcators and methods are used to assess both the dfferences beteen the attaned obecte alues as ell as the dfferences beteen decson arables of the Pareto sets. Especally the nely defned ndcators to compare Pareto sets n the decson space appeared to be useful to prode nsght n the structure of the Pareto sets. Copyrght c 204 SERSC 7

18 ( Z ) =0; ( Z ) =/5 ( Z ) =2/5 ( Z ) =3/5 ( Z ) =4/5 ( Z ) = Fgure 6. The Attanment Plots of the Outcome of Optmzaton Processes -5. The Blue Surface s attaned by none of the Pareto Sets, hle the Dark Red Surface s attaned by all Pareto Sets The frst concluson from ths analyss s that the dfferences n the attaned obecte alues are not too bg among the runs: nether thn runs th the same parameters nor beteen runs th dfferent parameter settngs. Hoeer, the mutaton rate seems to be of nfluence on the result, so t seems se to try at least a fe dfferent alues for ths parameter hen settng up a ne case study. Secondly, based on the aerage dstance to the nearest soluton n the decson space and dfference beteen the fractons of nonzero decson arables n to dfferent Pareto sets, t can be concluded that the decsons to be taken are substantally dfferent n all analyzed pars of Pareto sets, so the adsed decsons are nfluenced by the randomness of the algorthm. When takng decsons based on Pareto sets that are constructed by such an algorthm, more knoledge s needed on the senstty of decson arables th respect to the obecte alues, to dentfy mportant and less mportant decson arables. Thrdly, aratons caused by dfferent parameters n the NSGA-II algorthm are slghtly hgher than the aratons caused by randomness thn the algorthm, based on seeral methods for comparson of Pareto sets. Hoeer, because ths dfference s too small to be substantal, e conclude that the nfluence of the parameter alues to be chosen n the algorthm s not bg, mplyng that the rsk for runnng the algorthm th bad parameter alues s small. Acknoledgements We thank The Netherlands Organsaton for Scentfc Research (NWO) for fundng ths research n the program Sustanable Accessblty of the Randstad (SAR): ths research s part of the proect Strategy toards Sustanable and Relable Multmodal Transport n the Randstad (SRMT). Furthermore, e thank the Cty Regon of Amsterdam for the use of the Venom data. 8 Copyrght c 204 SERSC

19 References [] C. Coello, S. de Computacón and C. Zacatenco, "Tenty years of eolutonary mult-obecte optmzaton: A hstorcal e of the feld", IEEE Computatonal Intellgence Magazne, ol., no., (2006). [2] L. J. J. Wsmans, E. C. an Berkum and M. C. J. Blemer, "Comparson of Multobecte Eolutonary Algorthms for Optmzaton of Externaltes by Usng Dynamc Traffc Management Measures", Transportaton Research Record: Journal of the Transportaton Research Board, ol. 2263, (20). [3] V. Guhare and J. K. Hao, "Transt netork desgn and schedulng: A global ree", Transportaton Research Part A: Polcy and Practce, ol. 42, no. 0, (2008). [4] H. Yang and M. Bell, "Models and algorthms for road netork desgn: a ree and some ne deelopments", Transport Rees, ol. 8, no. 3, (998). [5] E. Mandoabch, R. Farahan, W. Dullaert and W. Szeto, "Hybrd Eolutonary Metaheurstcs for Concurrent Mult-Obecte Desgn of Urban Road and Publc Transt Netorks", Netorks and Spatal Economcs, ol. 2, no. 3, (202). [6] Y. Hamdouch, M. Floran, D. Hearn and S. Laphongpanch, "Congeston prcng for mult-modal transportaton systems", Transportaton Research Part B: Methodologcal, ol. 4, no. 3, (2007). [7] A. Chen, J. Km, S. Lee and Y. Km, "Stochastc mult-obecte models for netork desgn problem", Expert Systems th Applcatons, ol. 37, no. 2, (200). [8] S. Sharma, S. Ukkusur and T. Mathe, "Pareto Optmal Multobecte Optmzaton for Robust Transportaton Netork Desgn Problem", Transportaton Research Record: Journal of the Transportaton Research Board 2090, (2009). [9] G. Cantarella and A. Vtetta, "The mult-crtera road netork desgn problem n an urban area", Transportaton, ol. 33, no. 6, (2006). [0] Y. Yn and S. Laphongpanch, "Internalzng emsson externalty on road netorks", Transportaton Research Part D: Transport and Enronment, ol., no. 4, (2006). [] A. Sumalee, S. Shepherd and A. May, "Road user chargng desgn: dealng th mult-obectes and constrants", Transportaton, ol. 36, no. 2, (2009). [2] B. Beltran, S. Carrese, E. Cpran and M. Petrell, "Transt netork desgn th allocaton of green ehcles: A genetc algorthm approach", Transportaton Research Part C: Emergng Technologes, ol. 7, no. 5, (2009). [3] P. Chakroborty, "Genetc Algorthms for Optmal Urban Transt Netork Desgn", Computer-Aded Cl and Infrastructure Engneerng, ol. 8, no. 3, (2003). [4] E. Cpran, S. Gor and M. Petrell, "Transt netork desgn: A procedure and an applcaton to a large urban area", Transportaton Research Part C: Emergng Technologes, ol. 20, no., (202). [5] W. Fan and R. Machemehl, "Optmal Transt Route Netork Desgn Problem th Varable Transt Demand: Genetc Algorthm Approach", Journal of Transportaton Engneerng, ol. 32, no., (2006). [6] M. Gallo, B. Montella and L. D Acerno, "The transt netork desgn problem th elastc demand and nternalsaton of external costs: An applcaton to ral frequency optmsaton", Transportaton Research Part C: Emergng Technologes, ol. 9, no. 6, (20). [7] A. Mauttone and M. Urquhart, "A mult-obecte metaheurstc approach for the transt netork desgn problem", Publc Transport, ol., no. 4, (2009). [8] L. dell'olo, J. Moura and A. Ibeas, "B-Leel Mathematcal Programmng Model for Locatng Bus Stops and Optmzng Frequences", Transportaton Research Record: Journal of the Transportaton Research Board 97, (2006). [9] S. Tahmasseby, "Relablty n urban publc transport netork assessment and desgn", TU Delft, (2009). [20] F. Vt, S. F. Catalano, M. L, C. Lndeld and H. J. an Zuylen, "Optmzaton Problem th Dynamc Route-Departure Tme Choce and Prcng", 82nd Annual Meetng of the Transportaton Research Board, Washngton, D.C., (2003) January 2-6. [2] Y. Sheff, "Urban transportaton netorks: Equlbrum analyss th mathematcal programmng methods", Prentce-Hall, Engleood Clffs, Ne Jersey, (985). [22] Z. Gao, J. Wu and H. Sun, "Soluton algorthm for the b-leel dscrete netork desgn problem", Transportaton Research Part B: Methodologcal, ol. 39, no. 6, (2005). [23] K. Deb, "Mult obecte Optmzaton Usng Eolutonary Algorthms", John Wley & Sons, Chchester, UK (200). [24] K. Deb, A. Pratap, S. Agaral and T. Meyaran, "A fast and eltst multobecte genetc algorthm: NSGA-II", IEEE transactons on eolutonary computaton, ol. 6, no. 2, (2002). [25] E. Ztzler, J. Knoles and L. Thele, "Qualty Assessment of Pareto Set Approxmatons, n Multobecte Optmzaton", Edted J. Branke, K. Deb, K. Mettnen and R. Słońsk, Sprnger Berln Hedelberg, (2008). [26] L. Whle, P. Hngston, L. Barone and S. Huband, "A faster algorthm for calculatng hyperolume, Eolutonary Computaton", IEEE Transactons, ol. 0, no., (2006). Copyrght c 204 SERSC 9

20 [27] E. Ztzler and L. Thele, "Multobecte eolutonary algorthms: A comparate case study and the strength pareto approach", IEEE transactons on eolutonary computaton, no. 3, no. 4, (999). [28] J. Luttk, "The alue of trees, ater and open space as reflected by house prces n the Netherlands", Landscape and Urban Plannng, ol. 48, no. 3-4, (2000). [29] INFRAS, "Artems: Assessment and relablty of transport emsson models and nentory systems", Road Emsson Model Model Descrpton. Workpackage 00. Delerable, no. 3, INFRAS, (2007). Authors Tes Brands, s a PhD canddate at the Centre for Transport studes at the Unersty of Tente (The Netherlands). He receed hs BSc degree (Cl Engneerng) n 2004 and hs to MSc degrees (Cl Engneerng and Appled Mathematcs) n Hs MSc thess as on optmzaton of dynamc road prcng and as conducted at the Dutch consultancy company Goudappel Coffeng. After graduaton, Tes started orkng at Goudappel Coffeng as a publc transport consultant, th specalzaton n publc transport modelng and data analyss. Parallel to hs ork as a consultant, he started a PhD proect at the Centre for Transport Studes on mult-obecte optmzaton of multmodal passenger transportaton netorks. Erc an Berkum n the early 80 s Erc an Berkum receed hs BSc and MSc n Appled Mathematcs at the Unersty of Tente (The Netherlands). After a relately short career as a softare engneer at Cap Gemn he moed n 988 to Goudappel Coffeng, a traffc and transport consultancy company n Deenter. He orked there for 2 years n hch he has had seeral postons n ths company. Hs man nterest has been on transport modelng, netorks, trael behaour and traffc management. In 998 he became a parttme professor n traffc management at the Centre for Transport Studes. In 2009 he left Goudappel Coffeng and became a full professor and also head of the Centre for Transport Studes at the Unersty of Tente. 20 Copyrght c 204 SERSC

Integer Programming Formulations for the Uncapacitated Vehicle Routing p-hub Center Problem

Integer Programming Formulations for the Uncapacitated Vehicle Routing p-hub Center Problem 21st Internatonal Congress on Modellng and Smulaton, Gold Coast, Australa, 29 No to 4 Dec 2015 www.mssanz.org.au/modsm2015 Integer Programmng Formulatons for the Uncapactated Vehcle Routng p-hub Center

More information

SCHEDULING OF CONSTRUCTION PROJECTS BY MEANS OF EVOLUTIONARY ALGORITHMS

SCHEDULING OF CONSTRUCTION PROJECTS BY MEANS OF EVOLUTIONARY ALGORITHMS SCHEDULING OF CONSTRUCTION PROJECTS BY MEANS OF EVOLUTIONARY ALGORITHMS Magdalena Rogalska 1, Wocech Bożeko 2,Zdzsław Heduck 3, 1 Lubln Unversty of Technology, 2- Lubln, Nadbystrzycka 4., Poland. E-mal:rogalska@akropols.pol.lubln.pl

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

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

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

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

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

Using Multi-objective Metaheuristics to Solve the Software Project Scheduling Problem

Using Multi-objective Metaheuristics to Solve the Software Project Scheduling Problem Usng Mult-obectve Metaheurstcs to Solve the Software Proect Schedulng Problem Francsco Chcano Unversty of Málaga, Span chcano@lcc.uma.es Francsco Luna Unversty of Málaga, Span flv@lcc.uma.es Enrque Alba

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

行 政 院 國 家 科 學 委 員 會 補 助 專 題 研 究 計 畫 成 果 報 告 期 中 進 度 報 告

行 政 院 國 家 科 學 委 員 會 補 助 專 題 研 究 計 畫 成 果 報 告 期 中 進 度 報 告 行 政 院 國 家 科 學 委 員 會 補 助 專 題 研 究 計 畫 成 果 報 告 期 中 進 度 報 告 畫 類 別 : 個 別 型 計 畫 半 導 體 產 業 大 型 廠 房 之 設 施 規 劃 計 畫 編 號 :NSC 96-2628-E-009-026-MY3 執 行 期 間 : 2007 年 8 月 1 日 至 2010 年 7 月 31 日 計 畫 主 持 人 : 巫 木 誠 共 同

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

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

Formulating & Solving Integer Problems Chapter 11 289

Formulating & Solving Integer Problems Chapter 11 289 Formulatng & Solvng Integer Problems Chapter 11 289 The Optonal Stop TSP If we drop the requrement that every stop must be vsted, we then get the optonal stop TSP. Ths mght correspond to a ob sequencng

More information

Conversion between the vector and raster data structures using Fuzzy Geographical Entities

Conversion between the vector and raster data structures using Fuzzy Geographical Entities Converson between the vector and raster data structures usng Fuzzy Geographcal Enttes Cdála Fonte Department of Mathematcs Faculty of Scences and Technology Unversty of Combra, Apartado 38, 3 454 Combra,

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

+ + + - - This circuit than can be reduced to a planar circuit

+ + + - - This circuit than can be reduced to a planar circuit MeshCurrent Method The meshcurrent s analog of the nodeoltage method. We sole for a new set of arables, mesh currents, that automatcally satsfy KCLs. As such, meshcurrent method reduces crcut soluton to

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

Ants Can Schedule Software Projects

Ants Can Schedule Software Projects Ants Can Schedule Software Proects Broderck Crawford 1,2, Rcardo Soto 1,3, Frankln Johnson 4, and Erc Monfroy 5 1 Pontfca Unversdad Católca de Valparaíso, Chle FrstName.Name@ucv.cl 2 Unversdad Fns Terrae,

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

Credit Limit Optimization (CLO) for Credit Cards

Credit Limit Optimization (CLO) for Credit Cards Credt Lmt Optmzaton (CLO) for Credt Cards Vay S. Desa CSCC IX, Ednburgh September 8, 2005 Copyrght 2003, SAS Insttute Inc. All rghts reserved. SAS Propretary Agenda Background Tradtonal approaches to credt

More information

Lecture 2: Single Layer Perceptrons Kevin Swingler

Lecture 2: Single Layer Perceptrons Kevin Swingler Lecture 2: Sngle Layer Perceptrons Kevn Sngler kms@cs.str.ac.uk Recap: McCulloch-Ptts Neuron Ths vastly smplfed model of real neurons s also knon as a Threshold Logc Unt: W 2 A Y 3 n W n. A set of synapses

More information

Optimization of network mesh topologies and link capacities for congestion relief

Optimization of network mesh topologies and link capacities for congestion relief Optmzaton of networ mesh topologes and ln capactes for congeston relef D. de Vllers * J.M. Hattngh School of Computer-, Statstcal- and Mathematcal Scences Potchefstroom Unversty for CHE * E-mal: rwddv@pu.ac.za

More information

Activity Scheduling for Cost-Time Investment Optimization in Project Management

Activity Scheduling for Cost-Time Investment Optimization in Project Management PROJECT MANAGEMENT 4 th Internatonal Conference on Industral Engneerng and Industral Management XIV Congreso de Ingenería de Organzacón Donosta- San Sebastán, September 8 th -10 th 010 Actvty Schedulng

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

Simulation and optimization of supply chains: alternative or complementary approaches?

Simulation and optimization of supply chains: alternative or complementary approaches? Smulaton and optmzaton of supply chans: alternatve or complementary approaches? Chrstan Almeder Margaretha Preusser Rchard F. Hartl Orgnally publshed n: OR Spectrum (2009) 31:95 119 DOI 10.1007/s00291-007-0118-z

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

Preventive Maintenance and Replacement Scheduling: Models and Algorithms

Preventive Maintenance and Replacement Scheduling: Models and Algorithms Preventve Mantenance and Replacement Schedulng: Models and Algorthms By Kamran S. Moghaddam B.S. Unversty of Tehran 200 M.S. Tehran Polytechnc 2003 A Dssertaton Proposal Submtted to the Faculty of the

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

A DYNAMIC CRASHING METHOD FOR PROJECT MANAGEMENT USING SIMULATION-BASED OPTIMIZATION. Michael E. Kuhl Radhamés A. Tolentino-Peña

A DYNAMIC CRASHING METHOD FOR PROJECT MANAGEMENT USING SIMULATION-BASED OPTIMIZATION. Michael E. Kuhl Radhamés A. Tolentino-Peña Proceedngs of the 2008 Wnter Smulaton Conference S. J. Mason, R. R. Hll, L. Mönch, O. Rose, T. Jefferson, J. W. Fowler eds. A DYNAMIC CRASHING METHOD FOR PROJECT MANAGEMENT USING SIMULATION-BASED OPTIMIZATION

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

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

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

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

IT09 - Identity Management Policy

IT09 - Identity Management Policy IT09 - Identty Management Polcy Introducton 1 The Unersty needs to manage dentty accounts for all users of the Unersty s electronc systems and ensure that users hae an approprate leel of access to these

More information

Sciences Shenyang, Shenyang, China.

Sciences Shenyang, Shenyang, China. Advanced Materals Research Vols. 314-316 (2011) pp 1315-1320 (2011) Trans Tech Publcatons, Swtzerland do:10.4028/www.scentfc.net/amr.314-316.1315 Solvng the Two-Obectve Shop Schedulng Problem n MTO Manufacturng

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

"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

2008/8. An integrated model for warehouse and inventory planning. Géraldine Strack and Yves Pochet

2008/8. An integrated model for warehouse and inventory planning. Géraldine Strack and Yves Pochet 2008/8 An ntegrated model for warehouse and nventory plannng Géraldne Strack and Yves Pochet CORE Voe du Roman Pays 34 B-1348 Louvan-la-Neuve, Belgum. Tel (32 10) 47 43 04 Fax (32 10) 47 43 01 E-mal: corestat-lbrary@uclouvan.be

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

Popularity Based Adaptive Content Delivery Scheme with In-Network Caching

Popularity Based Adaptive Content Delivery Scheme with In-Network Caching Popularty Based Adapte Content Delery Scheme wth In-Network Cachng Jeong Yun Km, Gyu Myoung Lee, and Jun Kyun Cho To sole the ncreasng popularty of deo streamng serces oer the Internet, recent research

More information

FORECASTING TELECOMMUNICATION NEW SERVICE DEMAND BY ANALOGY METHOD AND COMBINED FORECAST

FORECASTING TELECOMMUNICATION NEW SERVICE DEMAND BY ANALOGY METHOD AND COMBINED FORECAST Yugoslav Journal of Operatons Research 5 (005), Number, 97-07 FORECAING ELECOMMUNICAION NEW ERVICE DEMAND BY ANALOGY MEHOD AND COMBINED FORECA Feng-Jenq LIN Department of Appled Economcs Natonal I-Lan

More information

Robust Design of Public Storage Warehouses. Yeming (Yale) Gong EMLYON Business School

Robust Design of Public Storage Warehouses. Yeming (Yale) Gong EMLYON Business School Robust Desgn of Publc Storage Warehouses Yemng (Yale) Gong EMLYON Busness School Rene de Koster Rotterdam school of management, Erasmus Unversty Abstract We apply robust optmzaton and revenue management

More information

Realistic Image Synthesis

Realistic Image Synthesis Realstc Image Synthess - Combned Samplng and Path Tracng - Phlpp Slusallek Karol Myszkowsk Vncent Pegoraro Overvew: Today Combned Samplng (Multple Importance Samplng) Renderng and Measurng Equaton Random

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

Forecasting the Demand of Emergency Supplies: Based on the CBR Theory and BP Neural Network

Forecasting the Demand of Emergency Supplies: Based on the CBR Theory and BP Neural Network 700 Proceedngs of the 8th Internatonal Conference on Innovaton & Management Forecastng the Demand of Emergency Supples: Based on the CBR Theory and BP Neural Network Fu Deqang, Lu Yun, L Changbng School

More information

Dynamic Fleet Management for Cybercars

Dynamic Fleet Management for Cybercars Proceedngs of the IEEE ITSC 2006 2006 IEEE Intellgent Transportaton Systems Conference Toronto, Canada, September 17-20, 2006 TC7.5 Dynamc Fleet Management for Cybercars Fenghu. Wang, Mng. Yang, Ruqng.

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 efficient constraint handling methodology for multi-objective evolutionary algorithms

An efficient constraint handling methodology for multi-objective evolutionary algorithms Rev. Fac. Ing. Unv. Antoqua N. 49. pp. 141-150. Septembre, 009 An effcent constrant handlng methodology for mult-objectve evolutonary algorthms Una metodología efcente para manejo de restrccones en algortmos

More information

Efficient Project Portfolio as a tool for Enterprise Risk Management

Efficient Project Portfolio as a tool for Enterprise Risk Management Effcent Proect Portfolo as a tool for Enterprse Rsk Management Valentn O. Nkonov Ural State Techncal Unversty Growth Traectory Consultng Company January 5, 27 Effcent Proect Portfolo as a tool for Enterprse

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

Nathalie Perrier Bruno Agard Pierre Baptiste Jean-Marc Frayret André Langevin Robert Pellerin Diane Riopel Martin Trépanier.

Nathalie Perrier Bruno Agard Pierre Baptiste Jean-Marc Frayret André Langevin Robert Pellerin Diane Riopel Martin Trépanier. A Survey of Models and Algorthms for Emergency Response Logstcs n Electrc Dstrbuton Systems - Part I: Relablty Plannng wth Fault Consderatons Nathale Perrer Bruno Agard Perre Baptste Jean-Marc Frayret

More information

A GENETIC ALGORITHM-BASED METHOD FOR CREATING IMPARTIAL WORK SCHEDULES FOR NURSES

A GENETIC ALGORITHM-BASED METHOD FOR CREATING IMPARTIAL WORK SCHEDULES FOR NURSES 82 Internatonal Journal of Electronc Busness Management, Vol. 0, No. 3, pp. 82-93 (202) A GENETIC ALGORITHM-BASED METHOD FOR CREATING IMPARTIAL WORK SCHEDULES FOR NURSES Feng-Cheng Yang * and We-Tng Wu

More information

Global Optimization Algorithms with Application to Non-Life Insurance

Global Optimization Algorithms with Application to Non-Life Insurance Global Optmzaton Algorthms wth Applcaton to Non-Lfe Insurance Problems Ralf Kellner Workng Paper Char for Insurance Economcs Fredrch-Alexander-Unversty of Erlangen-Nürnberg Verson: June 202 GLOBAL OPTIMIZATION

More information

Availability-Based Path Selection and Network Vulnerability Assessment

Availability-Based Path Selection and Network Vulnerability Assessment Avalablty-Based Path Selecton and Network Vulnerablty Assessment Song Yang, Stojan Trajanovsk and Fernando A. Kupers Delft Unversty of Technology, The Netherlands {S.Yang, S.Trajanovsk, F.A.Kupers}@tudelft.nl

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

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

A method for a robust optimization of joint product and supply chain design

A method for a robust optimization of joint product and supply chain design DOI 10.1007/s10845-014-0908-5 A method for a robust optmzaton of jont product and supply chan desgn Bertrand Baud-Lavgne Samuel Bassetto Bruno Agard Receved: 10 September 2013 / Accepted: 21 March 2014

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

A hybrid global optimization algorithm based on parallel chaos optimization and outlook algorithm

A hybrid global optimization algorithm based on parallel chaos optimization and outlook algorithm Avalable onlne www.ocpr.com Journal of Chemcal and Pharmaceutcal Research, 2014, 6(7):1884-1889 Research Artcle ISSN : 0975-7384 CODEN(USA) : JCPRC5 A hybrd global optmzaton algorthm based on parallel

More information

APPLICATION OF PROBE DATA COLLECTED VIA INFRARED BEACONS TO TRAFFIC MANEGEMENT

APPLICATION OF PROBE DATA COLLECTED VIA INFRARED BEACONS TO TRAFFIC MANEGEMENT APPLICATION OF PROBE DATA COLLECTED VIA INFRARED BEACONS TO TRAFFIC MANEGEMENT Toshhko Oda (1), Kochro Iwaoka (2) (1), (2) Infrastructure Systems Busness Unt, Panasonc System Networks Co., Ltd. Saedo-cho

More information

The Stochastic Guaranteed Service Model with Recourse for Multi-Echelon Warehouse Management

The Stochastic Guaranteed Service Model with Recourse for Multi-Echelon Warehouse Management The Stochastc Guaranteed Servce Model wth Recourse for Mult-Echelon Warehouse Management Jörg Rambau, Konrad Schade 1 Lehrstuhl für Wrtschaftsmathematk Unverstät Bayreuth Bayreuth, Germany Abstract The

More information

A New Task Scheduling Algorithm Based on Improved Genetic Algorithm

A New Task Scheduling Algorithm Based on Improved Genetic Algorithm A New Task Schedulng Algorthm Based on Improved Genetc Algorthm n Cloud Computng Envronment Congcong Xong, Long Feng, Lxan Chen A New Task Schedulng Algorthm Based on Improved Genetc Algorthm n Cloud Computng

More information

Meta-Analysis of Hazard Ratios

Meta-Analysis of Hazard Ratios NCSS Statstcal Softare Chapter 458 Meta-Analyss of Hazard Ratos Introducton Ths module performs a meta-analyss on a set of to-group, tme to event (survval), studes n hch some data may be censored. These

More information

Mooring Pattern Optimization using Genetic Algorithms

Mooring Pattern Optimization using Genetic Algorithms 6th World Congresses of Structural and Multdscplnary Optmzaton Ro de Janero, 30 May - 03 June 005, Brazl Moorng Pattern Optmzaton usng Genetc Algorthms Alonso J. Juvnao Carbono, Ivan F. M. Menezes Luz

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

Energy Efficient Routing in Ad Hoc Disaster Recovery Networks

Energy Efficient Routing in Ad Hoc Disaster Recovery Networks Energy Effcent Routng n Ad Hoc Dsaster Recovery Networks Gl Zussman and Adran Segall Department of Electrcal Engneerng Technon Israel Insttute of Technology Hafa 32000, Israel {glz@tx, segall@ee}.technon.ac.l

More information

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

Copyright. Satish V S K Ukkusuri

Copyright. Satish V S K Ukkusuri Copyrght by Satsh V S K Ukkusur 2005 The Dssertaton Commttee for Satsh V S K Ukkusur certfes that ths s the approved verson of the followng dssertaton: Accountng for Uncertanty, Robustness and Onlne Informaton

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

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

Dynamic Scheduling of Emergency Department Resources

Dynamic Scheduling of Emergency Department Resources Dynamc Schedulng of Emergency Department Resources Junchao Xao Laboratory for Internet Software Technologes, Insttute of Software, Chnese Academy of Scences P.O.Box 8718, No. 4 South Fourth Street, Zhong

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

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

Statistical Methods to Develop Rating Models

Statistical Methods to Develop Rating Models Statstcal Methods to Develop Ratng Models [Evelyn Hayden and Danel Porath, Österrechsche Natonalbank and Unversty of Appled Scences at Manz] Source: The Basel II Rsk Parameters Estmaton, Valdaton, and

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

) 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

A DATA MINING APPLICATION IN A STUDENT DATABASE

A DATA MINING APPLICATION IN A STUDENT DATABASE JOURNAL OF AERONAUTICS AND SPACE TECHNOLOGIES JULY 005 VOLUME NUMBER (53-57) A DATA MINING APPLICATION IN A STUDENT DATABASE Şenol Zafer ERDOĞAN Maltepe Ünversty Faculty of Engneerng Büyükbakkalköy-Istanbul

More information

IMPACT ANALYSIS OF A CELLULAR PHONE

IMPACT ANALYSIS OF A CELLULAR PHONE 4 th ASA & μeta Internatonal Conference IMPACT AALYSIS OF A CELLULAR PHOE We Lu, 2 Hongy L Bejng FEAonlne Engneerng Co.,Ltd. Bejng, Chna ABSTRACT Drop test smulaton plays an mportant role n nvestgatng

More information

Network Security Situation Evaluation Method for Distributed Denial of Service

Network Security Situation Evaluation Method for Distributed Denial of Service Network Securty Stuaton Evaluaton Method for Dstrbuted Denal of Servce Jn Q,2, Cu YMn,2, Huang MnHuan,2, Kuang XaoHu,2, TangHong,2 ) Scence and Technology on Informaton System Securty Laboratory, Bejng,

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

Vehicle Routing Problem with Time Windows for Reducing Fuel Consumption

Vehicle Routing Problem with Time Windows for Reducing Fuel Consumption 3020 JOURNAL OF COMPUTERS, VOL. 7, NO. 12, DECEMBER 2012 Vehcle Routng Problem wth Tme Wndows for Reducng Fuel Consumpton Jn L School of Computer and Informaton Engneerng, Zhejang Gongshang Unversty, Hangzhou,

More information

Inter-Ing 2007. INTERDISCIPLINARITY IN ENGINEERING SCIENTIFIC INTERNATIONAL CONFERENCE, TG. MUREŞ ROMÂNIA, 15-16 November 2007.

Inter-Ing 2007. INTERDISCIPLINARITY IN ENGINEERING SCIENTIFIC INTERNATIONAL CONFERENCE, TG. MUREŞ ROMÂNIA, 15-16 November 2007. Inter-Ing 2007 INTERDISCIPLINARITY IN ENGINEERING SCIENTIFIC INTERNATIONAL CONFERENCE, TG. MUREŞ ROMÂNIA, 15-16 November 2007. UNCERTAINTY REGION SIMULATION FOR A SERIAL ROBOT STRUCTURE MARIUS SEBASTIAN

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

NONLINEAR OPTIMIZATION FOR PROJECT SCHEDULING AND RESOURCE ALLOCATION UNDER UNCERTAINTY

NONLINEAR OPTIMIZATION FOR PROJECT SCHEDULING AND RESOURCE ALLOCATION UNDER UNCERTAINTY NONLINEAR OPTIMIZATION FOR PROJECT SCHEDULING AND RESOURCE ALLOCATION UNDER UNCERTAINTY A Dssertaton Presented to the Faculty of the Graduate School of Cornell Unversty In Partal Fulfllment of the Requrements

More information

Business Process Improvement using Multi-objective Optimisation K. Vergidis 1, A. Tiwari 1 and B. Majeed 2

Business Process Improvement using Multi-objective Optimisation K. Vergidis 1, A. Tiwari 1 and B. Majeed 2 Busness Process Improvement usng Mult-objectve Optmsaton K. Vergds 1, A. Twar 1 and B. Majeed 2 1 Manufacturng Department, School of Industral and Manufacturng Scence, Cranfeld Unversty, Cranfeld, MK43

More information

Automated information technology for ionosphere monitoring of low-orbit navigation satellite signals

Automated information technology for ionosphere monitoring of low-orbit navigation satellite signals Automated nformaton technology for onosphere montorng of low-orbt navgaton satellte sgnals Alexander Romanov, Sergey Trusov and Alexey Romanov Federal State Untary Enterprse Russan Insttute of Space Devce

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

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

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

INVESTIGATION OF VEHICULAR USERS FAIRNESS IN CDMA-HDR NETWORKS

INVESTIGATION OF VEHICULAR USERS FAIRNESS IN CDMA-HDR NETWORKS 21 22 September 2007, BULGARIA 119 Proceedngs of the Internatonal Conference on Informaton Technologes (InfoTech-2007) 21 st 22 nd September 2007, Bulgara vol. 2 INVESTIGATION OF VEHICULAR USERS FAIRNESS

More information

Enterprise Master Patient Index

Enterprise Master Patient Index Enterprse Master Patent Index Healthcare data are captured n many dfferent settngs such as hosptals, clncs, labs, and physcan offces. Accordng to a report by the CDC, patents n the Unted States made an

More information

Dynamic Routing in Self-Healing MPLS Networks

Dynamic Routing in Self-Healing MPLS Networks Dynamc Rng n Self-Healng MPS Networs Krzysztof Walowa Char of Systems and Computer Networs, Faculty of Electroncs, Wroclaw Unersty of Technology, Wybrzeze Wyspansego 27, 50-370 Wroclaw, Poland walowa@zss.pwr.wroc.pl

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

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

Scatter search approach for solving a home care nurses routing and scheduling problem

Scatter search approach for solving a home care nurses routing and scheduling problem Scatter search approach for solvng a home care nurses routng and schedulng problem Bouazza Elbenan 1, Jacques A. Ferland 2 and Vvane Gascon 3* 1 Département de mathématque et nformatque, Faculté des scences,

More information

Fragility Based Rehabilitation Decision Analysis

Fragility Based Rehabilitation Decision Analysis .171. Fraglty Based Rehabltaton Decson Analyss Cagdas Kafal Graduate Student, School of Cvl and Envronmental Engneerng, Cornell Unversty Research Supervsor: rcea Grgoru, Professor Summary A method s presented

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

SOLVING CARDINALITY CONSTRAINED PORTFOLIO OPTIMIZATION PROBLEM BY BINARY PARTICLE SWARM OPTIMIZATION ALGORITHM

SOLVING CARDINALITY CONSTRAINED PORTFOLIO OPTIMIZATION PROBLEM BY BINARY PARTICLE SWARM OPTIMIZATION ALGORITHM SOLVIG CARDIALITY COSTRAIED PORTFOLIO OPTIMIZATIO PROBLEM BY BIARY PARTICLE SWARM OPTIMIZATIO ALGORITHM Aleš Kresta Klíčová slova: optmalzace portfola, bnární algortmus rojení částc Key words: portfolo

More information

Optimal Choice of Random Variables in D-ITG Traffic Generating Tool using Evolutionary Algorithms

Optimal Choice of Random Variables in D-ITG Traffic Generating Tool using Evolutionary Algorithms Optmal Choce of Random Varables n D-ITG Traffc Generatng Tool usng Evolutonary Algorthms M. R. Mosav* (C.A.), F. Farab* and S. Karam* Abstract: Impressve development of computer networks has been requred

More information

Cross-Selling in a Call Center with a Heterogeneous Customer Population

Cross-Selling in a Call Center with a Heterogeneous Customer Population OPERATIONS RESEARCH Vol. 57, No. 2, March Aprl 29, pp. 299 313 ssn 3-364X essn 1526-5463 9 572 299 nforms do 1.1287/opre.18.568 29 INFORMS Cross-Sellng n a Call Center wth a Heterogeneous Customer Populaton

More information