A Probabilistic Approach to Latent Cluster Analysis

Size: px
Start display at page:

Download "A Probabilistic Approach to Latent Cluster Analysis"

Transcription

1 Procdngs of th Twnty-Thrd Intrnatonal Jont Confrnc on Artfcal Intllgnc A Probablstc Approach to Latnt Clstr Analyss Zhpng X R Dong, Zhnghng Dng, Zhnyng H, Wdong Yang School of Comptr Scnc Fdan Unrsty, Shangha Chna {xzp, , , zhnyng, wdyang}@fdan.d.cn Abstract Facng a larg nmbr of clstrng soltons, clstr nsmbl mthod prods an ffct approach to aggrgatng thm nto a bttr on. In ths papr, w propos a nol clstr nsmbl mthod from probablstc prspct. It assms that ach clstrng solton s gnratd from a latnt clstr modl, ndr th control of two probablstc paramtrs. Ths, th clstr nsmbl problm s rformlatd nto an optmzaton problm of maxmm lklhood. An EM-styl algorthm s dsgnd to sol ths problm. It can dtrmn th nmbr of clstrs atomatcally. Exprmnal rslts ha shown that th proposd algorthm otprforms th stat-of-th-art mthods ncldng EAC-AL,,, and. Frthrmor, t has bn shown that or algorthm s stabl n th prdctd nmbrs of clstrs. 1 Introdcton Th goal of clstr analyss s to dscor th ndrlyng strctr of a datast (Jan t al., 1999; Jan, It normally parttons a st of obcts so that th obcts wthn th sam grop ar smlar whl thos from dffrnt grops ar dssmlar. A larg nmbr of clstrng algorthms ha bn proposd,.g. k-mans, Spctral Clstrng, Hrarchcal Clstrng, Slf-Organzng Maps, to nam bt a fw, yt no sngl on s abl to sccssflly ach ths goal for all datasts. On th sam data, dffrnt algorthms, or n mltpl rns of th sam algorthm wth dffrnt paramtrs, oftn lad to clstrng soltons that ar dstnct from ach othr. Confrontd wth a larg nmbr of clstrng soltons, clstr nsmbl or clstrng aggrgaton mthods ha mrgd, whch try to combn dffrnt clstrng soltons nto a consnss on, n ordr to mpro th qalty of componnt clstrng soltons (Vga-Pons and Rz-Shlclopr, Clstr nsmbl mthods sally consst of two or thr phass: th nsmbl gnraton phas to prodc a arty of clstrng soltons; thn th nsmbl slcton phas to slct a sbst of ths clstrng soltons, whch s optonal; and fnally th consnss phas to ndc a nfd partton by combnng th componnt ons. In th gnraton phas, dffrnt clstrng soltons can b gnratd by dffrnt clstrng algorthms, th sam algorthm wth dffrnt paramtr sttngs or ntalzaton, and ncton of random dstrbanc nto data st sch as data rsamplng (Mna-Bdgol t al., 2004, random procton (Frn and Brodly, 2003, and random fatr slcton (Strhl and Ghosh, Followng th gnraton phas, an optonal nmbl slcton phas wll slct or prn ths clstrng soltons accordng to thr qalts and drsts (Frn and Ln, 2008; Azm and Frn, In ths papr, w focs on th fnal phas - clstrng combnaton. Thr ar a lot of algorthms for th combnaton, whch can b catgorzd accordng to th knd of nformaton xplotd. Th algorthm proposd hr falls nto th catgory makng s of th parws smlarts btwn obcts, whch form a co-assocaton matrx n th contxt of clstr nsmbls. Any clstrng algorthm can b appld on ths nw smlarty matrx to fnd a consnss partton. Ednc Accmmlaton Clstrng (Frd and Jan, 2005, or EAC n short, prforms a hrarchcal clstrng of arag lnkag (AL or sngl lnkag (SL on co-assocatong matrx, whr a maxmm lftm crtron s proposd to dtrmn th nmbr of clstrs. Clstr-basd Smlarty Parttonng ( algorthm (Strhl and Ghosh, 2002 ss a graph-partonng algorthm nstad, bt rqrs th nmbr of clstrs b spcfd manally. Anothr algorthm (Strhl and Ghosh, 2002 can b thoght as an approxmaton to. Ot of ths catgory, algorthm maks a clstrng of clstrs basd on th smlarts btwn clstrs, and thn assgns obcts to ts closst mta-clstr. For a thorogh lst of rlatd algorthms, plas rfr to th sry papr by Vga-Pons and Rz-Shlclopr (2011. Althogh ths mthods ha achd som sccss, thy ar stll dfcnt n sral aspcts: frst, thy lack of thortc ndrpnnng; scond, thy thnk that all th clstrng soltons b of th sam qalty, and ths assgn th sam wght to ach clstrng solton; last bt not th last, most of thm (xcpt EAC rqr th nmbr of clstrs to b spcfd manally. As to th maxmm lftm crtron adoptd by EAC algorthm, t s mor or lss a rl-of-thmb that s lack of stfcaton. As w shall s n th xprmnts, th maxmm lftm crtron s nstabl n dtrmnng th nmbr of clstrs. 1813

2 To tackl ths problms, w propos a probablstc mthod calld LAtnt Clstr Analyss, or LACA n short. It assms that thr s a latnt clstr modl whch s nobsrabl. All th obsrd Clstrng soltons ar gnratd from th latnt modl ndr th control of two probablstc paramtrs. Or obct s to sk th latnt clstr modl wth th maxmm lklhood. Ths papr s organzd as follows. In Scton 2, w ntrodc th latnt clstr modl and bld ts conncton wth th obsrd clstrng soltons. W dot Scton 3 to an EM-styl algorthm for nfrrng th latnt clstr modl from th obsrd clstrng soltons. In Scton 4, w prsnt th xprmntal rslts of th proposd mthod compard wth sral stat-of-th-art clstr nsmbl algorthms. Fnally, w mak th conclson n Scton 5. 2 Latnt Clstr Modl Lt X = {x 1 2, n } b a st of n obcts, whr ach obct x may b rprsntd as a mltdmnsonal ctor, a strng, or n any othr form. Takng X as npt, a clstrng algorthm (calld clstrr prodcs a clstrng solton that parttons th n obcts nto grops. By rnnng clstrng algorthms mltpl tms, w can obsr an nsmbl of clstrng soltons, E = {C 1, C 2,, C E }, whr ach clstrng solton C (1 E parttons th n obcts nto grops c, c,, c. C Wth rspct to a gn clstrng solton C, a co-assocaton rlatonshp btwn two obcts x and x s dfnd accordng to whthr thy ar assgnd to th sam grop: 1, f ck C sch that { x } c k. (1 0, othrws Two obcts x and x ar sad to b co-assocatd n th clstrng solton C f 1 ; othrws, thy ar not co-assocatd. Gn th nsmbl of obsrd clstrng solton, what w wold lk to xplor s th latnt clstr modl. W dnot th latnt clstr modl as = { 1, 2,, s }, whr l (1 l s s a clstr rprsntd as a sbst of obcts, and s s th ral nmbr of clstrs. In ths papr, w assm that ths s latnt clstrs ar non-orlappng,.., = for 1 s. Basd on th latnt clstr modl, w dfn th co-clstr fncton for a gn par of obcts {x, x }:, f k sch that { x } k (2, othrws Ths latnt clstr modl srs as th maor factor n dtrmnng what clstrng soltons can b obsrd, whl othr factors sch as th bas of appld clstrng algorthm nflncs th obsrd rslts and lads to som fals posts and fals ngats. To bld th conncton btwn a clstrng solton C and th latnt clstr modl, w ntrodc two probablstc paramtrs: Th paramtr to dnot th condtonal probablty that two obcts ar co-assocatd n C gn that thy ar co-clstr n th hddn modl, that s = Pr( =1 =1; and Th paramtr r to dnot th condtonal probablty that two obcts ar co-assocatd n C gn that thy ar not co-clstr n th hddn modl, that s r = Pr( =1 =0. Inttly, ach obsrd clstrng solton prods som dncs abot th latnt (nobsrabl co-clstr rlatonshp btwn obcts. Gn a st E of clstrng soltons, or obct s to maxmz th postror probablty of th latnt clstr modl, as follows: * E Pr( arg max Pr( E arg max (3 Pr( E whr E = Pr(E s th lklhood fncton. Assm that ths clstrng soltons ar ndpndnt of ach othr and pror probablts of all th possbl latnt clstr modls ar dstrbtd nformly, t can b wrttn as th followng maxmm-lklhood problm: * arg maxl ( E arg max C, (4 whr C = Pr(C s th lklhood of th latnt modl gn th obsrd clstrng solton C (or th probablty of obsrng C gn th latnt modl Th dnc of obsrng a clstrng solton C can b dcomposd nto th dnc st of co-assocaton rlatonshps of all obct pars,.. whthr a par of obcts {x } ar assgnd to th sam grop. Thn, w ha C ( r ( r (5 : : : : Takng logrthm on both sds, w gt th log-lklhood: L ( C log C (6,,,, n11 log n log( n01 log r n log( r whr, n1 {{ x } and } rprsnts th nmbr of obct pars that ar co-clstr n th hddn modl and also co-assocatd n C ;, n10 {{ x } and 0} rprsnts th nmbr of obct pars that ar co-clstr n th hddn modl, bt not co-assocatd n th obsrd clstrng rslt C ;, n01 {{ x } 0 and 1} rprsnts th nmbr of obct pars that ar not co-clstr n th hddn modl, bt co-assocatd n th obsrd clstrng rslt C ; and 1814

3 , n {{ x } 0 and } rprsnts th 00 nmbr of th obct pars that ar not co-clstr n also not co-assocatd n th clstrng rslt C. By sbstttng qatons (5 or (6 nto (4, w can rformlat th optmzaton problm as: * arg max l ( C arg max L( C (7 3 Algorthm Dsgn Unfortnatly, th latnt clstr modl and th probablstc paramtrs of ths obsrd clstrng rslts ar all nknown, whch mak t mpossbl to sk th solton drctly. Hr, w proposd an EM-styl algorthm to dal wth th problm, dpctd n Fgr 1. Fgr 1. Th flowchart of LACA algorthm Th algorthm conssts of for maor stps: Stp 1 (Paramtr Intalzaton: ntalz th probablstc paramtrs for ach clstrng solton; Stp 2 (Latnt Modl Gnraton: fxng th probablstc paramtrs, look for a nar-optmal solton (a latnt modl to th maxmm-lklhood problm wth a hll clmbng stratgy; Stp 3 (Paramtr Estmaton: fxng th latnt clstr modl, stmat th probablstc paramtrs for ach clstrng solton; Stp 4 (Conrgnc Tst: Rpat Stp 2 and Stp 3 ntl conrgnc. 3.1 Paramtr Intalzaton Gn E obsrd clstrng soltons C 1, C 2,, C E, w s cont(x to dnot th nmbr of clstrng soltons whr th obcts x and x ar co-assocatd, and rcont(x to dnot th nmbr of clstrng soltons whr x and x ar not co-assocatd, that s: and Clstrng 1 Clstrng 2... Clstrng E Paramtr Intalzat cont(x = rcont(x = Latnt Modl Gnraton Paramtr Rstmaton Conrgnc? (8 (. (9 It s dnt that cont(x + rcont(x = E for 1 n,. So far as paramtr ntalzaton s concrnd, t s takn for grantd that dffrnt clstrng soltons b qally No Ys Otpt th Solton plasbl. Th hghr s th al of cont(x, th mor probably th two obcts x and x ar co-clstrd. Hnc, th two probablstc paramtrs for ach clstrng solton C s ntalzd as th followng M-stmats: cont( x x ESS, :, (10 cont( x ESS and r : 1 rcont( x ESS, (11 rcont( x, y ESS whr ESS, standng for qalnt sampl sz, s st as 30 by dfalt. 3.2 Latnt Modl Gnraton Onc th probablstc paramtrs ar ntalzd or r-stmatd, th latnt clstr modl can dtrmnd n a hll-clmbng mannr wth rspct to al th log lklhood fncton. Lt s start wth an mpty latnt modl whr ach obct corrsponds to a snglton clstr, and ths any two obcts ar not co-clstr. Thn, w tratly mrg two clstrs nto a largr on. Th slcton crtron for mrgng at ach stp s dscrbd blow, stp by stp. Lt = { 1, 2,, t } b th latnt clstr modl at th crrnt stp, and ( b prodcd by mrgng and n,,. D to th fact that ( n (,, n = ( n n and 11, (,, ( n ( n = ( n, n, t can b drd that: ( E E (12 (,, ( n11 n11 log ( n r Frthr, t s dnt that: and ( n ( n,, 11 n11 11 (, x n (,, 10 n10 ( x Sbstttng (13 and (14 nto (12, w gt: ( E E x log r x 01, log 01 r, (13 ( log. (14 r (15 By ntrchangng th ordr of smmaton, t can b drd that: ( E E (16 log ( log x r r If w dfn th affnty scor btwn two obcts x and x otd by a clstrng solton C as: 1815

4 scor x x ( log ( log, (17 r r w can sm p all th scors otd by C 1, C 2,, C E nto th corrspondng ntry M[ ] of a scor matrx M, that s, M [ ] scor ( x. (18 Sbstttng (17 and (18 nto (16, w ha: ( E E M[ ] x (19 Howr, sng qaton (19 drctly as th slcton crtron may faor mrgng largr clstrs or smallr ons. Ths, w choos to mrg two clstrs s and t sch that ( E E ( s, t arg max, (20 1 arg max M[ ], x If w thnk of th scor matrx M as a smlarty matrx, th slcton crtron s actally th arag-lnkag (AL. As a rslt, a hrarchcal clstrng wth arag lnkag can b appld on th matrx M. What s mportant s that th lmnts n th scor matrx has clar probablstc manng: ach lmnt rprsnts actally th log-lklhood rato of th corrpondng two obcts bng co-clstr to bng not. Hddn Modl Infrnc a Hrarchcal clstrng By fxng all th paramtrs and r (1 E, th scor matrx M can b constrctd accordng to (17 and (18. Th hddn modl can b gnratd by applyng th followng agglommrat hrarchcal clstrng on M: Stp 1: W start from th smplst snglton modl 0 whr ach clstr conssts of on and only on obct. Stp 2: Lt t dnot th crrnt traton, and st t = 1. Stp 3: At ach traton t, lt b th clstr modl n th pros traton,.., = t1. Wthot loss of gnralty, w dnot ={ 1, 2,, }. Two clstrs s and t n ar slctd accordng to qaton (20. Stp 4: If th arag lnk btwn s and t s ngat thn trmnat th loop and otpt as th gnratd hddn clstr modl; othrws, contn to stp 5. Stp 5: Th clstrs slctd at stp 3 gt mrgd nto a nw clstr nw = s t. W pdat by rmong s and t, and nsrtng nw. Th pdatd thn srs as th clstr modl n th crrnt traton, that s t =. Stp 6: If only two clstrs ar lft, trmnat and otpt ; othrws, st t = t+1, and go to stp 3. Stoppng Crtron: Ths procss s rpatd ntl w can not fnd a par of clstrs wth post arag lnk (Stp 4, or thr ar only two clstrs lft (Stp Paramtr R-stmaton Onc a latnt clstr modl s gnratd, t can b sd to stmat th probablstc paramtrs and r for ach clstrng solton C. Snc th paramtr rprsnts th probablty that two obcts ar co-assocatd n C on condton that thy ar co-clstr n th latnt modl, that s, r = Pr( =1 =1, t can b stmatd as: {( } 0. ESS, (21 {( } ESS whr th ESS s also st as 30 by dfalt. Bcas th paramtr r dnots th probablty that two obcts ar co-assocatd n C on condton that thy ar not co-clstr n, that s r = Pr( =1 =0, t can b stmatd as: {( }. ESS r. (22 {( } ESS If a clstrng solton assgns ach obct nto a dstnct grop, th corrspondng and r wll b both clos to 0. On th othr xtrm, f t assgns all obcts nto a sngl grop, th corrspondng and r wll b both nar to Conrgnc Tst Onc th probablstc paramtrs and r for ach clstrng solton C ar r-stmatd, w compt th dffrnc btwn th r-stmatd als and th pros ons. If th sm of absolt dffrncs or all clstrng soltons s lss than a sr-spcfd thrshold al, w consdr th algorthm as conrgd and otpt th latnt modl. 4 Exprmnts W ha condctd xtns xprmnts to compar LACA wth sral stat-of-th-art clstr nsmbl mthods. Or xprmnts ar dsgnd to dmonstrat: 1 LACA s mor stabl than EAC-AL n dtrmnng th nmbr of clstrs; 2 LACA otprforms EAC-AL whch s also abl to dtrmnng th nmbr of clstrs atomatcally; 3 A arant rson of LACA, calld, otprforms, and. 4.1 Exprmntal Sttngs DataSt #Obct #Fatr #Class Irs Glass Ecol Lbras Sgmntaton Sd Pma Pndgts Tabl 1: Dscrptons of th datasts. 1816

5 Data sts. W s ght data sts from th UCI machn larnng rpostory (Frank and Asncon, 2010 n or xprmnts. Th charactrstcs of th data sts ar smmarzd n Tabl 1. Not that, for Pndgts, w randomly slct 100 obcts from ach class. Clstr Ensmbl Gnraton. W choos to s th K-mans algorthm (MacQn, 1967 as or bas clstrr, bcas of ts poplarty n many pros clstr nsmbl stds. At ach rn, w gnrat a clstr nsmbl of 200 clstrng soltons for a gn data st. To b mor spcfc, for a datast of n obcts and m fatrs, ach clstrng solton s prodcd as follows: Th sz s of fatr sbst s frstly dtrmnd by randomly drawng an ntgr al from th rang [mns, maxs], whr mns s st to b 3, and maxs s st to b m. A random fatr sbst FS of sz s s gnratd by drawng s dffrnt fatrs from th orgnal m fatrs. An random ntgr al K s drawn from [mnk, maxk], whr mnk s st to b 2, and maxk s st to b n/15. A clstrng solton s obtand by applyng K-mans algorthm on th datast, wth accss to all th obcts, bt only th s fatrs n FS. Ealaton Crtron. As all th datasts ar labld, w s th class labls as a srrogat for th tr ndrlyng strctr of th data. Two commonly sd masrs, Normalzd Mtal Informaton ( and F-masrs, ar chosn to alat or approach aganst othrs. (Strhl and Ghosh, 2002 trats clstr labls X and class labls Y as random arabls and maks a tradoff btwn th mtal nformaton and th nmbr of clstrs: I( X, Y, H ( X H ( Y whr I( s th mtal nformaton mtrc and H( s th ntropy mtrc. F-masr (Mannng t al., 2008 ws a clstrng solton (on a datast wth n obcts as a srs of n(n1/2 dcsons, on for ach par of obcts. It maks a compoms btwn th prcson and th rcall of ths dcsons: Prcson Rcall F. Prcson Rcall 4.2 Stablty of Prdctd Clstr Nmbrs To th bst of or knowldg, most clstr nsmbl mthods rly on a sr-spcfd nmbr of clstrs. Th only xcpton s th maxmm lftm crtron sd n EAC-AL mthod. In ordr to stdy th stablty of or algorthm and EAC-AL n prdctd nmbr of clstrs, w gnrat 30 clstr nsmbls for ach datast n th way as dscrbd n Scton 4.1. Or algorthm and EAC-AL ar appld on ths clstr nsmbls to gt thr prdctd clstr nmbrs. Th statstcs abot ths nmbrs ar prsntd n Tabl 2. It can b sn from th tabl that th rang [Mn, Max] of EAC-AL s mch wdr than that of LACA on ach datast, sggstng that th clstr nmbrs prdctd by EAC-AL flctats a lot for ach data st, and spcally for Pndgts and Pma. Smlar obsratons can also b mad from th standard daton of clstr nmbrs on ach datast. W conctr that ths s bcas lf tm s not always ffct n th prdcton of clstr nmbrs, bcas th maxmm lftm stratgy s mor or lss a rl of thmb, lack of thortc stfcaton. As to th arag al of th prdctd clstr nmbrs, LACA s largr than EAC-AL on 3 datasts, and smallr on 5, showng nconsstncy n som dgr. Datast Mthod Mn Max Arag Std D Irs LACA EAC-AL Glass LACA EAC-AL Ecol LACA EAC-AL Lbras LACA EAC-AL Sgmntaton LACA EAC-AL Sd LACA EAC-AL Pma LACA EAC-AL Pndgts LACA EAC-AL Tabl 2: Statstcs of prdctd clstr nmbrs 4.3 Comparson wth EAC-AL Tabl 3 rports th and F-masr als of or algorthm and EAC-AL on th sam clstr nsmbls. Each al rportd hr s obtand by aragng across 30 rns. W can s that or algorthm prforms bttr than EAC-AL on 7 ot of 8 datasts. Th only xcpton s on th Lbras datast. W conctr that t s bcas th arag clstr nmbr prdctd by EAC-AL s closr to th ral nmbr of classs n Lbras. F-masr Datast LACA EAC-AL LACA EAC-AL Irs Glass Ecol Lbras Sgmntaton Sd Pma Pndgts Tabl 3: F-masr and als of LACA and EAC-AL 4.4 Comparson wth, and 1817

6 rs glass col lbras clstr nmbr k sgmntaton clstr nmbr k clstr nmbr k sd clstr nmbr k clstr nmbr k pma clstr nmbr k Fgr 2. Comparson of,, and clstr nmbr k pndgts clstr nmbr k F al F al clstr nmbr k sgmntaton rs clstr nmbr k clstr nmbr k sd D to th fact that most xstng clstr nsmbl mthods rqr a sr-spcfd nmbr of clstrs, to mak a far comparson wth thm, w mak a small modfcaton of LACA by accptng a sr-spcfd clstr nmbr k, rsltng n a arant rson calld. In, whn th probablstc paramtrs gt conrgd, w forc th hrarchcal clstrng to stop mrgng only f thr ar xactly k clstrs lft, whch ar thn sd as th consnss clstrng of. On ach datast of l classs, w compar wth,, and by aryng th clstr nmbr k from l to 3 l wth stp sz 1. Fgr 2 and Fgr 3 dpct th s and th F-masrs of,, and, rspctly, whch ar also argad or 30 rns, wth dffrnt sr-spcfd clstr nmbrs. It s obos that th cr of or mthod s bttr or at last comptt on almost all th datasts. Th only xcpton s obsrd on th Pma datast, whr th of or mthod s lowr than th othrs. Bsds, w also fnd that th cr of or mthod s mor smooth and stabl across ths dffrnt k als. Ths sggsts that or mthod has achd hgh qalts consstntly on ths lls of hrarchcal clstrng. F al F al glass clstr nmbr k F al F al col clstr nmbr k pma clstr nmbr k Fgr 3. F-masr comparson of,, and F al F al clstr nmbr k pndgts Conclsons In ths papr, w proposd a nol clstr nsmbl approach by assmng that th obsrd clstrng soltons ar gnratd from a latnt clstr modl. An EM-styl algorthm, calld LACA, was dsgnd and mplmntd to maxmz th lklhood fncton. It has xhbtd a satsfactory prformanc on th xprmntal datasts, for two rasons: frstly, t can mak a stabl and rlabl prdcton of th clstr nmbrs; scondly, ot of ach bas clstrng solton s wghtd whch rflcts th qalty of th bas solton. Acknowldgmnts Ths work s spportd by Natonal Arplan Rsarch Program (MJ-Y , Natonal Natral Scnc Fnd of Chna (No , Shangha Hgh-Tch Proct (11-43 and Shangha Ladng Acadmc Dscpln Proct (No. B114. W ar gratfl to th anonymos rwrs for thr alabl commnts. lbras clstr nmbr k 1818

7 Rfrncs [Azm and Frn, 2009] Jaad Azm Xaol Z. Frn. Adapt clstr nsmbl slcton. In Procdngs of th 21 st Intrnatonal Jont Confrnc on Artfcal Intllgnc, pags , [Frn and Brodly, 2003] Xaol Z. Frn, and Carla E. Brodly. Random procton for hgh dmnsonal data clstrng: a clstr nsmbl approach. In Procdngs of th 20 th Intrnatonal Confrnc on Machn Larnng, pags , [Frn and Ln, 2008] Xaol Z. Frn, and W Ln. Clstr nsmbl slcton. Statstcal Analyss and Data Mnng, 1(3: , 2008 [Frank and Asncon, 2010] A. Frank, and A. Asncon. UCI Machn Larnng Rpostory. Irn, CA: Unrsty of Calforna, School of Informaton and Comptr Scnc, [ [Frd and Jan, 2005] Ana L.N. Frd, and Anl K. Jan. Combnng mltpl clstrngs sng dnc accmlaton. IEEE Transactons on Pattrn Rcognton and Machn Intllgnc. 27(6: , [Jan t al., 1999] Anl K. Jan, M.N. Mrty, P.J. Flynn. Data clstrng: a rw. ACM Comptng Srys, 31(3: , [Jan, 2010] Anl K. Jan. Data clstrng: 50 yars byond K-mans. Pattrn Rcognton Lttrs, 31(8: , [MacQn, 1967] J. MacQn. Som mthods for classfcatons and analyss of mltarat obsratons. In Procdngs of th Ffth Brkly Symposm on Mathmatcs, Statstcs and Probablty, Unrsty of Calforna Prss, pags , [Mannng t al., 2008] Chrstophr D. Mannng, Prabhakar Raghaan, and Hnrch Schtz. Introdcton to Informaton Rtral, Cambrdg UnrstyPrss, [Mna-Bdgol t al., 2004] Bhroz Mna-Bdgol Alxandr Topchy, Wllam F. Pnch. Ensmbls of parttons a data rsamplng. In Procdngs of Intrnatonal Confrnc on Informaton Tchnology: Codng and Comptng (ITCC 2004, pags , [Strhl and Ghosh, 2002] Alxandr Strhl, and Joydp Ghosh. Clstr nsmbls - a knowldg rs framwork for combnng mltpl parttons. Jornal of Machn Larnng Rsarch, 3: , [Vga-Pons and Rz-Shlclopr, 2011] Sandro Vga-Pons, and Jos Rz-Shlclopr. A srry of clstrng nsmbl algorthms. Intrnatonal Jornal of Pattrn Rcognton and Artfcal Intllgnc, 25(3: ,

Reliability-Driven Reputation Based Scheduling for Public-Resource Computing Using GA

Reliability-Driven Reputation Based Scheduling for Public-Resource Computing Using GA 2009 Intrnatonal Confrnc on Advancd Informaton Ntworkng and Applcatons Rlablty-Drvn Rputaton Basd Schdulng for Publc-Rsourc Computng Usng GA Xaofng Wang #, Ch Shn Yo*, Rakumar Buyya* 2, Jnshu Su # 2 #Collg

More information

Modern Portfolio Theory (MPT) Statistics

Modern Portfolio Theory (MPT) Statistics Modrn Portfolo Thory (MPT) Statstcs Mornngstar Mthodology Papr Novmr 30, 007 007 Mornngstar, Inc. All rghts rsrvd. Th nformaton n ths documnt s th proprty of Mornngstar, Inc. Rproducton or transcrpton

More information

Online Load Balancing and Correlated Randomness

Online Load Balancing and Correlated Randomness Onln Load Balancng and Corrlatd Randomnss Sharayu Moharr, Sujay Sanghav Wrlss Ntworng and Communcatons Group (WNCG) Dpartmnt of Elctrcal & Computr Engnrng Th Unvrsty of Txas at Austn Austn, TX 787, USA

More information

QUANTITATIVE METHODS CLASSES WEEK SEVEN

QUANTITATIVE METHODS CLASSES WEEK SEVEN QUANTITATIVE METHODS CLASSES WEEK SEVEN Th rgrssion modls studid in prvious classs assum that th rspons variabl is quantitativ. Oftn, howvr, w wish to study social procsss that lad to two diffrnt outcoms.

More information

ANALYSIS OF ORDER-UP-TO-LEVEL INVENTORY SYSTEMS WITH COMPOUND POISSON DEMAND

ANALYSIS OF ORDER-UP-TO-LEVEL INVENTORY SYSTEMS WITH COMPOUND POISSON DEMAND 8 th Intrnatonal Confrnc of Modlng and Smulaton - MOSIM - May -2, 2 - Hammamt - Tunsa Evaluaton and optmzaton of nnovatv producton systms of goods and srvcs ANALYSIS OF ORDER-UP-TO-LEVEL INVENTORY SYSTEMS

More information

How To Write A Recipe Card

How To Write A Recipe Card Employng Rlvanc Fdback to mbd Contnt and Srvc Importanc nto th Slcton Procss of Compost Cloud Srvcs Dmosthns Kyrazs, Nkolaos Doulams, Gorg Kousours, Andras nychtas, arnos Thmstoclous, Vasslos C. Vscouks

More information

Life Analysis for the Main bearing of Aircraft Engines

Life Analysis for the Main bearing of Aircraft Engines f Analyss for th Man barng of Arcraft Engns Png n a, Xaolng Zhang a, png H a, anglang Dng a a School of Mchancs, Elctronc, and Industral Engnrng, Unvrsty of Elctronc Scnc and Tchnology of Chna, Chngdu,

More information

Authenticated Encryption. Jeremy, Paul, Ken, and Mike

Authenticated Encryption. Jeremy, Paul, Ken, and Mike uthntcatd Encrypton Jrmy Paul Kn and M Objctvs Examn thr mthods of authntcatd ncrypton and dtrmn th bst soluton consdrng prformanc and scurty Basc Componnts Mssag uthntcaton Cod + Symmtrc Encrypton Both

More information

Mininum Vertex Cover in Generalized Random Graphs with Power Law Degree Distribution

Mininum Vertex Cover in Generalized Random Graphs with Power Law Degree Distribution Mnnum Vrtx Covr n Gnralzd Random Graphs wth Powr Law Dgr Dstrbuton André L Vgnatt a, Murlo V G da Slva b a DINF Fdral Unvrsty of Paraná Curtba, Brazl b DAINF Fdral Unvrsty of Tchnology - Paraná Curtba,

More information

The example is taken from Sect. 1.2 of Vol. 1 of the CPN book.

The example is taken from Sect. 1.2 of Vol. 1 of the CPN book. Rsourc Allocation Abstract This is a small toy xampl which is wll-suitd as a first introduction to Cnts. Th CN modl is dscribd in grat dtail, xplaining th basic concpts of C-nts. Hnc, it can b rad by popl

More information

The influence of advertising on the purchase of pharmaceutical products

The influence of advertising on the purchase of pharmaceutical products Th nflunc of advrtsng on th purchas of pharmacutcal products Jana VALEČKOVÁ, VŠB-TU Ostrava Abstract Th sz of th pharmacutcal markt and pharmacutcal sals s ncrasng constantly. Th markt s floodd wth nw

More information

Section 3: Logistic Regression

Section 3: Logistic Regression Scton 3: Logstc Rgrsson As our motvaton for logstc rgrsson, w wll consdr th Challngr dsastr, th sx of turtls, collg math placmnt, crdt card scorng, and markt sgmntaton. Th Challngr Dsastr On January 28,

More information

Protecting E-Commerce Systems From Online Fraud

Protecting E-Commerce Systems From Online Fraud Protctng E-Commrc Systms From Onln Fraud Frst Author P.PhanAlkhya Studnt, Dpartmnt of Computr Scnc and Engnrng, QIS Collg of Engnrng & Tchnology, ongol, Andhra Pradsh, Inda. Scond Author Sk.Mahaboob Basha

More information

PRACTICAL ADVANTAGES OF USING THE MECHANICS OF CONTINUUM TO ANALYSE DEFORMATIONS OBTAINED FROM GEODETIC SURVEY

PRACTICAL ADVANTAGES OF USING THE MECHANICS OF CONTINUUM TO ANALYSE DEFORMATIONS OBTAINED FROM GEODETIC SURVEY PRACTICAL ADVANTAGES OF USING THE MECHANICS OF CONTINUUM TO ANALYSE DEFORMATIONS OBTAINED FROM GEODETIC SURVEY Mlan TALICH Rsarch Insttut of Godsy, Topography and Cartography, Zdby 98, CZ-5 66, Czch Rpublc

More information

An RSA-based (t, n) threshold proxy signature scheme with freewill identities

An RSA-based (t, n) threshold proxy signature scheme with freewill identities Int. J. Informaton an Computr Scurty, Vol. 1, No. 1/2, 27 21 An RSA-bas (t, n) thrshol proxy sgnatur schm wth frwll ntts Ya-Fn Chang Grauat Insttut of Accountng, Natonal Chung Hsng Unvrsty, Tachung 42,

More information

ERLANG C FORMULA AND ITS USE IN THE CALL CENTERS

ERLANG C FORMULA AND ITS USE IN THE CALL CENTERS IFORTIO D OUITIO TEHOLOGIES D SERVIES, VOL. 9, O., RH 2 7 ERLG FORUL D ITS USE I THE LL ETERS Er HROY., Tbor ISUTH., atj KVKY. Dpartmnt of Tlcommuncatons, Faculty of Elctrcal Engnrng and Informaton Tchnology,

More information

Game of Platforms: Strategic Expansion into Rival (Online) Territory

Game of Platforms: Strategic Expansion into Rival (Online) Territory Gam of Platforms: Stratgc Expanson nto Rval (Onln) Trrtory Sagt Bar-Gll Ϯ Abstract Onln platforms, such as Googl, Facbook, or Amazon, ar constantly xpandng thr actvts, whl ncrasng th ovrlap n thr srvc

More information

Control of Perceived Quality of Service in Multimedia Retrieval Services: Prediction-based mechanism vs. compensation buffers

Control of Perceived Quality of Service in Multimedia Retrieval Services: Prediction-based mechanism vs. compensation buffers 1 Control of Prcvd Qualty of Srvc n ultmda Rtrval Srvcs: Prdcton-basd mchansm vs. compnsaton buffrs Aurlo La Cort, Alfo Lombardo, Srgo Palazzo, Govann Schmbra Isttuto d Informatca Tlcomuncazon, Unvrsty

More information

Personalized Web Search by User Interest Hierarchy

Personalized Web Search by User Interest Hierarchy Prsonalzd Wb arch by Usr Intrst Hrarchy Abstract Most of th wb sarch ngns ar dsgnd to srv all usrs, ndpndnt of th nds of any ndvdual usr. Prsonalzaton of wb sarch s to carry out rtrval for ach usr ncorporatng

More information

5 2 index. e e. Prime numbers. Prime factors and factor trees. Powers. worked example 10. base. power

5 2 index. e e. Prime numbers. Prime factors and factor trees. Powers. worked example 10. base. power Prim numbrs W giv spcial nams to numbrs dpnding on how many factors thy hav. A prim numbr has xactly two factors: itslf and 1. A composit numbr has mor than two factors. 1 is a spcial numbr nithr prim

More information

Bank Incentives, Economic Specialization, and Financial Crises in Emerging Economies

Bank Incentives, Economic Specialization, and Financial Crises in Emerging Economies Bank Incntvs, Economc Spcalzaton, and nancal Crss n Emrgng Economs Amar Gand a,*, Kos John b, and mma W. Snbt c a Cox School of Busnss, Southrn Mthodst Unvrsty, 6 Bshop Blvd., Dallas, TX 7575 USA b Strn

More information

GIBBS ENSEMBLE AND SOCKETON ESSENTIAL COMPONENTS FOR CREATING CLOUD COMPUTING

GIBBS ENSEMBLE AND SOCKETON ESSENTIAL COMPONENTS FOR CREATING CLOUD COMPUTING Intrnatonal Journal on Cloud Computng: Srvcs and Archtctur (IJCCSA),Vol.3, o.3, Jun 013 GIBBS SMBL AD SOCKTO SSTIAL COMPOTS FOR CRATIG CLOUD COMPUTIG I ITRT SYSTM Xaoquan Gao Rchrch t Dévloppmnt Drcton

More information

Constrained Renewable Resource Allocation in Fuzzy Metagraphs via Min- Slack

Constrained Renewable Resource Allocation in Fuzzy Metagraphs via Min- Slack Intrntonl Journl of ppld Oprtonl Rsrch Vol 1, No 1, pp 7-17, Summr 011 Journl hompg: wwworlur Constrnd Rnwl Rsourc llocton n Fuzzy Mtgrphs v Mn- Slck S S Hshmn* Rcvd: Jnury 31, 011 ; ccptd: My 1, 011 strct

More information

Term Structure of Interest Rates: The Theories

Term Structure of Interest Rates: The Theories Handou 03 Econ 333 Abdul Munasb Trm Srucur of Inrs Ras: Th Thors Trm Srucur Facs Lookng a Fgur, w obsrv wo rm srucur facs Fac : Inrs ras for dffrn maurs nd o mov oghr ovr m Fac : Ylds on shor-rm bond mor

More information

Adverse Selection and Moral Hazard in a Model With 2 States of the World

Adverse Selection and Moral Hazard in a Model With 2 States of the World Advrs Slction and Moral Hazard in a Modl With 2 Stats of th World A modl of a risky situation with two discrt stats of th world has th advantag that it can b natly rprsntd using indiffrnc curv diagrams,

More information

Advantageous Selection versus Adverse Selection in Life Insurance Market

Advantageous Selection versus Adverse Selection in Life Insurance Market Covr Pag Advantagous Slcton vrsus Advrs Slcton n f Insuranc Markt Ghadr Mahdav mahdav@conomcs.mbo.mda.kyoto-u.ac.j Post Doctoral Rsarch Assocat: Jaan Socty for th Promoton of Scnc JSPS, Graduat School

More information

No 28 Xianning West Road, Xi an No 70 Yuhua East Road, Shijiazhuang. yongchunliang@hotmail.com

No 28 Xianning West Road, Xi an No 70 Yuhua East Road, Shijiazhuang. yongchunliang@hotmail.com On-Ln Dynamc Cabl Ratng for Undrground Cabls basd on DTS and FEM Y.C.Lang *, Y.M.L School of Elctrcal Engnrng * Dpartmnt of Elctrcal and Informaton X an Jaotong Unvrsty Hb Unvrsty of Scnc and Tchnology

More information

Non-Linear and Unbalanced Three-Phase Load Static Compensation with Asymmetrical and Non Sinusoidal Supply

Non-Linear and Unbalanced Three-Phase Load Static Compensation with Asymmetrical and Non Sinusoidal Supply Non-Lnar and nbalancd Thr-Phas Load Statc Comnsaton wth Asymmtrcal and Non Snusodal Suly Rys S. Hrrra and P. Salmrón Elctrcal Engnrng Dartmnt Escula Poltécnca Suror, nvrsty of Hulva Ctra. Palos d la Frontra,

More information

TRACKING PERFORMANCE OF GPS RECEIVERS WITH MORE THAN ONE MULTIPATH

TRACKING PERFORMANCE OF GPS RECEIVERS WITH MORE THAN ONE MULTIPATH TRACKING ERFORANCE OF GS RECEIVERS WITH ORE THAN ONE ULTIATH ABSTRACT Chrstoph ACABIAU, Bnoît ROTURIER, Abdlahad BENHALLA CNS Rsarch Laboratory of th ENAC, ENAC, 7 avnu Edouard Bln, B 45, 3155 TOULOUSE

More information

Keywords Cloud Computing, Service level agreement, cloud provider, business level policies, performance objectives.

Keywords Cloud Computing, Service level agreement, cloud provider, business level policies, performance objectives. Volum 3, Issu 6, Jun 2013 ISSN: 2277 128X Intrnational Journal of Advancd Rsarch in Computr Scinc and Softwar Enginring Rsarch Papr Availabl onlin at: wwwijarcsscom Dynamic Ranking and Slction of Cloud

More information

Modelling Exogenous Variability in Cloud Deployments

Modelling Exogenous Variability in Cloud Deployments Modllng Exognous Varablty n Cloud Dploymnts Gulano Casal 1 Mrco Trbaston 2 g.casal@mpral.ac.u trbaston@pst.f.lmu.d 1 : Impral Collg London, London, Untd Kngdom 2 : Ludwg-Maxmlans-Unvrstät, Munch, Grmany

More information

Logistic Regression for Insured Mortality Experience Studies. Zhiwei Zhu, 1 Zhi Li 2

Logistic Regression for Insured Mortality Experience Studies. Zhiwei Zhu, 1 Zhi Li 2 Logstc Rgrsson for Insurd Mortalty Exprnc Studs Zhw Zhu, Zh L 2 Prsntd at th Lvng to 00 Symposum Orlando, Fla. January 8 0, 204 Copyrght 204 by th Socty of Actuars. All rghts rsrvd by th Socty of Actuars.

More information

Reputation Management for DHT-based Collaborative Environments *

Reputation Management for DHT-based Collaborative Environments * Rputaton Managmnt for DHT-basd Collaboratv Envronmnts * Natalya Fdotova, Luca Vltr Unvrsty of Parma, Italy Abstract Ths artcl addrsss a problm of ntgraton of rputaton managmnt mchansms and lookup procsss

More information

Buffer Management Method for Multiple Projects in the CCPM-MPL Representation

Buffer Management Method for Multiple Projects in the CCPM-MPL Representation Industra ngnrng & Managmnt Systms Vo No 4 Dcmbr 22 pp.397-45 ISSN 598-7248 ISSN 2234-6473 http://dx.do.org/.7232/ms.22..4.397 22 KII Buffr Managmnt Mthod for Mutp Projcts n th CCPM-MP Rprsntaton Nguyn

More information

Facts About Chronc Fatgu Syndrom - sample thereof

Facts About Chronc Fatgu Syndrom - sample thereof Cardac Toxcty n Chronc Fatgu Syndrom: Rsults from a Randomzd 40-Wk Multcntr Doubl-blnd Placbo Control Tral of Rntatolmod Bruc C. Stouch, Ph.D 1 Davd Strayr, M.D 2. Wllam Cartr, M.D 2. 1 Phladlpha Collg

More information

Rural and Remote Broadband Access: Issues and Solutions in Australia

Rural and Remote Broadband Access: Issues and Solutions in Australia Rural and Rmot Broadband Accss: Issus and Solutions in Australia Dr Tony Warrn Group Managr Rgulatory Stratgy Tlstra Corp Pag 1 Tlstra in confidnc Ovrviw Australia s gographical siz and population dnsity

More information

A Graph-based Proactive Fault Identification Approach in Computer Networks

A Graph-based Proactive Fault Identification Approach in Computer Networks A Graph-basd Proacti Fault Idntification Approach in Computr Ntworks Yijiao Yu, Qin Liu and Lianshng Tan * Dpartmnt of Computr Scinc, Cntral China Normal Unirsity, Wuhan 4379 PR China E-mail: yjyu, liuqin,

More information

EFFECT OF GEOMETRICAL PARAMETERS ON HEAT TRANSFER PERFORMACE OF RECTANGULAR CIRCUMFERENTIAL FINS

EFFECT OF GEOMETRICAL PARAMETERS ON HEAT TRANSFER PERFORMACE OF RECTANGULAR CIRCUMFERENTIAL FINS 25 Vol. 3 () January-March, pp.37-5/tripathi EFFECT OF GEOMETRICAL PARAMETERS ON HEAT TRANSFER PERFORMACE OF RECTANGULAR CIRCUMFERENTIAL FINS *Shilpa Tripathi Dpartmnt of Chmical Enginring, Indor Institut

More information

VOL. 25, NÚM. 54, EDICIÓN JUNIO 2007 PP. 122-155

VOL. 25, NÚM. 54, EDICIÓN JUNIO 2007 PP. 122-155 Ensayos sobr POÍTICA ECONÓMICA PENSION FUND MANAGERS AND THE STRUCTURE OF THE FOREIGN EXCHANGE MARKET HERNANDO VARGAS YANNETH ROCÍO BETANCOURT ENSAYOS SOBRE POÍTICA ECONÓMICA, VO. 25, NÚM. 54, EDICIÓN

More information

arijit_laha@infosys.com

arijit_laha@infosys.com art_laha@nfosys.com ABSRAC Enhancmnt of tchnology-basd systm support for knowldg workrs s an ssu of grat mportanc. h Knowldg work Support Systm (KwSS) framwork analyzs ths ssu from a holstc prspctv. KwSS

More information

Performance Evaluation

Performance Evaluation Prformanc Evaluation ( ) Contnts lists availabl at ScincDirct Prformanc Evaluation journal hompag: www.lsvir.com/locat/pva Modling Bay-lik rputation systms: Analysis, charactrization and insuranc mchanism

More information

erkeley / uc berkeley extension Be YoUR Best / be est with berkeley / uc berkeley With BerkELEY exten xtension / be your best with berkele

erkeley / uc berkeley extension Be YoUR Best / be est with berkeley / uc berkeley With BerkELEY exten xtension / be your best with berkele c rkley xtnsion yor st w rkly c rkley xtnsion st with rkly c rkley xtn Crtificat yor Program st in with rkly c rk xtnsion yor st with rkl c rkley xtnsion yor st w rkly Financial c Planning rkley xtnsion

More information

Traffic Information Estimation Methods Based on Cellular Network Data

Traffic Information Estimation Methods Based on Cellular Network Data TOPCO 崇 越 論 文 大 賞 論 文 題 目 ( 以 中 文 繕 寫 論 文 者 請 用 中 文 題 目 英 文 繕 寫 者 請 用 英 文 ): Traffc Informaton Estmaton Mthods Basd on Cuar Ntwork Data 報 名 編 號 : AI002 基 於 蜂 巢 網 路 資 料 之 交 通 資 訊 估 計 方 法 摘 要 近 年 來 隨 著 經

More information

A Note on Approximating. the Normal Distribution Function

A Note on Approximating. the Normal Distribution Function Applid Mathmatical Scincs, Vol, 00, no 9, 45-49 A Not on Approimating th Normal Distribution Function K M Aludaat and M T Alodat Dpartmnt of Statistics Yarmouk Univrsity, Jordan Aludaatkm@hotmailcom and

More information

C H A P T E R 1 Writing Reports with SAS

C H A P T E R 1 Writing Reports with SAS C H A P T E R 1 Writing Rports with SAS Prsnting information in a way that s undrstood by th audinc is fundamntally important to anyon s job. Onc you collct your data and undrstand its structur, you nd

More information

Lecture 20: Emitter Follower and Differential Amplifiers

Lecture 20: Emitter Follower and Differential Amplifiers Whits, EE 3 Lctur 0 Pag of 8 Lctur 0: Emittr Followr and Diffrntial Amplifirs Th nxt two amplifir circuits w will discuss ar ry important to lctrical nginring in gnral, and to th NorCal 40A spcifically.

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

CPS 220 Theory of Computation REGULAR LANGUAGES. Regular expressions

CPS 220 Theory of Computation REGULAR LANGUAGES. Regular expressions CPS 22 Thory of Computation REGULAR LANGUAGES Rgular xprssions Lik mathmatical xprssion (5+3) * 4. Rgular xprssion ar built using rgular oprations. (By th way, rgular xprssions show up in various languags:

More information

Traffic Flow Analysis (2)

Traffic Flow Analysis (2) Traffic Flow Analysis () Statistical Proprtis. Flow rat distributions. Hadway distributions. Spd distributions by Dr. Gang-Ln Chang, Profssor Dirctor of Traffic safty and Oprations Lab. Univrsity of Maryland,

More information

Managing the Outsourcing of Two-Level Service Processes: Literature Review and Integration

Managing the Outsourcing of Two-Level Service Processes: Literature Review and Integration Procdns of th 43rd Hawa Intrnatonal Confrnc on Systm Scncs - 2010 Manan th Outsourcn of Two-Lvl Srvc Procsss: Ltratur Rvw and Intraton Edal Pnkr Unvrsty of Rochstr pnkr@smon.rochstr.du Robrt Shumsky Dartmouth

More information

by John Donald, Lecturer, School of Accounting, Economics and Finance, Deakin University, Australia

by John Donald, Lecturer, School of Accounting, Economics and Finance, Deakin University, Australia Studnt Nots Cost Volum Profit Analysis by John Donald, Lcturr, School of Accounting, Economics and Financ, Dakin Univrsity, Australia As mntiond in th last st of Studnt Nots, th ability to catgoris costs

More information

Use a high-level conceptual data model (ER Model). Identify objects of interest (entities) and relationships between these objects

Use a high-level conceptual data model (ER Model). Identify objects of interest (entities) and relationships between these objects Chaptr 3: Entity Rlationship Modl Databas Dsign Procss Us a high-lvl concptual data modl (ER Modl). Idntify objcts of intrst (ntitis) and rlationships btwn ths objcts Idntify constraints (conditions) End

More information

Gold versus stock investment: An econometric analysis

Gold versus stock investment: An econometric analysis Intrnational Journal of Dvlopmnt and Sustainability Onlin ISSN: 268-8662 www.isdsnt.com/ijds Volum Numbr, Jun 202, Pag -7 ISDS Articl ID: IJDS20300 Gold vrsus stock invstmnt: An conomtric analysis Martin

More information

A Multi-Heuristic GA for Schedule Repair in Precast Plant Production

A Multi-Heuristic GA for Schedule Repair in Precast Plant Production From: ICAPS-03 Procdings. Copyright 2003, AAAI (www.aaai.org). All rights rsrvd. A Multi-Huristic GA for Schdul Rpair in Prcast Plant Production Wng-Tat Chan* and Tan Hng W** *Associat Profssor, Dpartmnt

More information

Upper Bounding the Price of Anarchy in Atomic Splittable Selfish Routing

Upper Bounding the Price of Anarchy in Atomic Splittable Selfish Routing Uppr Bounding th Pric of Anarchy in Atomic Splittabl Slfish Routing Kamyar Khodamoradi 1, Mhrdad Mahdavi, and Mohammad Ghodsi 3 1 Sharif Univrsity of Tchnology, Thran, Iran, khodamoradi@c.sharif.du Sharif

More information

FACULTY SALARIES FALL 2004. NKU CUPA Data Compared To Published National Data

FACULTY SALARIES FALL 2004. NKU CUPA Data Compared To Published National Data FACULTY SALARIES FALL 2004 NKU CUPA Data Compard To Publishd National Data May 2005 Fall 2004 NKU Faculty Salaris Compard To Fall 2004 Publishd CUPA Data In th fall 2004 Northrn Kntucky Univrsity was among

More information

Descriptive Models. Cluster Analysis. Example. General Applications of Clustering. Examples of Clustering Applications

Descriptive Models. Cluster Analysis. Example. General Applications of Clustering. Examples of Clustering Applications CMSC828G Prncples of Data Mnng Lecture #9 Today s Readng: HMS, chapter 9 Today s Lecture: Descrptve Modelng Clusterng Algorthms Descrptve Models model presents the man features of the data, a global summary

More information

PROXIMITY OPERATIONS OF ON-ORBIT SERVICING SPACECRAFT USING AN ECCENTRICITY/INCLINATION VECTOR SEPARATION

PROXIMITY OPERATIONS OF ON-ORBIT SERVICING SPACECRAFT USING AN ECCENTRICITY/INCLINATION VECTOR SEPARATION PROXMY OPERAONS O ON-ORB SERVCNG SPACECRA USNG AN ECCENRCYNCLNAON VECOR SEPARAON J. Sprmnn 1, S. D Amco 1 DLR, Grmn Spc Oprtons Cntr, 83 Wsslng, jorn.sprmnn@dlr.d DLR, Grmn Spc Oprtons Cntr, 83 Wsslng,

More information

WORKERS' COMPENSATION ANALYST, 1774 SENIOR WORKERS' COMPENSATION ANALYST, 1769

WORKERS' COMPENSATION ANALYST, 1774 SENIOR WORKERS' COMPENSATION ANALYST, 1769 08-16-85 WORKERS' COMPENSATION ANALYST, 1774 SENIOR WORKERS' COMPENSATION ANALYST, 1769 Summary of Dutis : Dtrmins City accptanc of workrs' compnsation cass for injurd mploys; authorizs appropriat tratmnt

More information

is knowing the car market inside out.

is knowing the car market inside out. Exprts s knowng th car markt nsd out. Vokswagn Group Lasng s both a ft managmnt company and a dvson of Vokswagn Fnanca Srvcs (UK) Lmtd. Ths aows us to offr a comprhnsv rang of srvcs wth packags to sut

More information

Incomplete 2-Port Vector Network Analyzer Calibration Methods

Incomplete 2-Port Vector Network Analyzer Calibration Methods Incomplt -Port Vctor Ntwork nalyzr Calibration Mthods. Hnz, N. Tmpon, G. Monastrios, H. ilva 4 RF Mtrology Laboratory Instituto Nacional d Tcnología Industrial (INTI) Bunos irs, rgntina ahnz@inti.gov.ar

More information

Entity-Relationship Model

Entity-Relationship Model Entity-Rlationship Modl Kuang-hua Chn Dpartmnt of Library and Information Scinc National Taiwan Univrsity A Company Databas Kps track of a company s mploys, dpartmnts and projcts Aftr th rquirmnts collction

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

Real-Time Evaluation of Email Campaign Performance

Real-Time Evaluation of Email Campaign Performance Singapor Managmnt Univrsity Institutional Knowldg at Singapor Managmnt Univrsity Rsarch Collction L Kong Chian School Of Businss L Kong Chian School of Businss 10-2008 Ral-Tim Evaluation of Email Campaign

More information

Question 3: How do you find the relative extrema of a function?

Question 3: How do you find the relative extrema of a function? ustion 3: How do you find th rlativ trma of a function? Th stratgy for tracking th sign of th drivativ is usful for mor than dtrmining whr a function is incrasing or dcrasing. It is also usful for locating

More information

Architecture of the proposed standard

Architecture of the proposed standard Architctur of th proposd standard Introduction Th goal of th nw standardisation projct is th dvlopmnt of a standard dscribing building srvics (.g.hvac) product catalogus basd on th xprincs mad with th

More information

TELL YOUR STORY WITH MYNEWSDESK The world's leading all-in-one brand newsroom and multimedia PR platform

TELL YOUR STORY WITH MYNEWSDESK The world's leading all-in-one brand newsroom and multimedia PR platform TELL YOUR STORY WITH MYNEWSDESK Th world's lading all-in-on brand nwsroom and multimdia PR platform SO WHAT'S THE STORY WITH MYNEWSDESK? Th world s lading all-in-on nwsroom and digital PR platform. Usd

More information

Basis risk. When speaking about forward or futures contracts, basis risk is the market

Basis risk. When speaking about forward or futures contracts, basis risk is the market Basis risk Whn spaking about forward or futurs contracts, basis risk is th markt risk mismatch btwn a position in th spot asst and th corrsponding futurs contract. Mor broadly spaking, basis risk (also

More information

Mathematics. Mathematics 3. hsn.uk.net. Higher HSN23000

Mathematics. Mathematics 3. hsn.uk.net. Higher HSN23000 hsn uknt Highr Mathmatics UNIT Mathmatics HSN000 This documnt was producd spcially for th HSNuknt wbsit, and w rquir that any copis or drivativ works attribut th work to Highr Still Nots For mor dtails

More information

Abstract. Introduction. Statistical Approach for Analyzing Cell Phone Handoff Behavior. Volume 3, Issue 1, 2009

Abstract. Introduction. Statistical Approach for Analyzing Cell Phone Handoff Behavior. Volume 3, Issue 1, 2009 Volum 3, Issu 1, 29 Statistical Approach for Analyzing Cll Phon Handoff Bhavior Shalini Saxna, Florida Atlantic Univrsity, Boca Raton, FL, shalinisaxna1@gmail.com Sad A. Rajput, Farquhar Collg of Arts

More information

An IAC Approach for Detecting Profile Cloning in Online Social Networks

An IAC Approach for Detecting Profile Cloning in Online Social Networks An IAC Approach for Dtcting Profil Cloning in Onlin Social Ntworks MortzaYousfi Kharaji 1 and FatmhSalhi Rizi 2 1 Dptartmnt of Computr and Information Tchnology Enginring,Mazandaran of Scinc and Tchnology,Babol,

More information

STATEMENT OF INSOLVENCY PRACTICE 3.2

STATEMENT OF INSOLVENCY PRACTICE 3.2 STATEMENT OF INSOLVENCY PRACTICE 3.2 COMPANY VOLUNTARY ARRANGEMENTS INTRODUCTION 1 A Company Voluntary Arrangmnt (CVA) is a statutory contract twn a company and its crditors undr which an insolvncy practitionr

More information

Sharp bounds for Sándor mean in terms of arithmetic, geometric and harmonic means

Sharp bounds for Sándor mean in terms of arithmetic, geometric and harmonic means Qian t al. Journal of Inqualitis and Applications (015) 015:1 DOI 10.1186/s1660-015-0741-1 R E S E A R C H Opn Accss Sharp bounds for Sándor man in trms of arithmtic, gomtric and harmonic mans Wi-Mao Qian

More information

RISK MANAGEMENT OF UNCERTAIN INNOVATION PROJECT BASED ON BAYESIAN RISK DECISION

RISK MANAGEMENT OF UNCERTAIN INNOVATION PROJECT BASED ON BAYESIAN RISK DECISION Sept 201. Vol., o. 2012-201 EAAS & ARF. All rghts reserve www.eaas-ornal.org ISS20-8269 RISK MAAGEMET OF UCERTAI IOVATIO PROJECT BASED O BAYESIA RISK DECISIO Yngchn Go College of Mathematcs & Compter Seence,

More information

Global Sourcing: lessons from lean companies to improve supply chain performances

Global Sourcing: lessons from lean companies to improve supply chain performances 3 rd Intrnational Confrnc on Industrial Enginring and Industrial Managmnt XIII Congrso d Ingniría d Organización Barclona-Trrassa, Sptmbr 2nd-4th 2009 Global Sourcing: lssons from lan companis to improv

More information

JOB-HOPPING IN THE SHADOW OF PATENT ENFORCEMENT

JOB-HOPPING IN THE SHADOW OF PATENT ENFORCEMENT JOB-HOPPING IN THE SHADOW OF PATENT ENFORCEMENT Rajshr Agarwal* Robrt H. Smth School of Busnss Unvrsty of Maryland rajshr@umd.du Martn Ganco Carlson School of Managmnt Unvrsty of Mnnsota mganco@umn.du

More information

METHODS FOR HANDLING TIED EVENTS IN THE COX PROPORTIONAL HAZARD MODEL

METHODS FOR HANDLING TIED EVENTS IN THE COX PROPORTIONAL HAZARD MODEL STUDIA OECONOMICA POSNANIENSIA 204, vol. 2, no. 2 (263 Jadwiga Borucka Warsaw School of Economics, Institut of Statistics and Dmography, Evnt History and Multilvl Analysis Unit jadwiga.borucka@gmail.com

More information

Constraint-Based Analysis of Gene Deletion in a Metabolic Network

Constraint-Based Analysis of Gene Deletion in a Metabolic Network Constraint-Basd Analysis of Gn Dltion in a Mtabolic Ntwork Abdlhalim Larhlimi and Alxandr Bockmayr DFG-Rsarch Cntr Mathon, FB Mathmatik und Informatik, Fri Univrsität Brlin, Arnimall, 3, 14195 Brlin, Grmany

More information

Planning and Managing Copper Cable Maintenance through Cost- Benefit Modeling

Planning and Managing Copper Cable Maintenance through Cost- Benefit Modeling Planning and Managing Coppr Cabl Maintnanc through Cost- Bnfit Modling Jason W. Rup U S WEST Advancd Tchnologis Bouldr Ky Words: Maintnanc, Managmnt Stratgy, Rhabilitation, Cost-bnfit Analysis, Rliability

More information

Cloud and Big Data Summer School, Stockholm, Aug., 2015 Jeffrey D. Ullman

Cloud and Big Data Summer School, Stockholm, Aug., 2015 Jeffrey D. Ullman Cloud and Big Data Summr Scool, Stockolm, Aug., 2015 Jffry D. Ullman Givn a st of points, wit a notion of distanc btwn points, group t points into som numbr of clustrs, so tat mmbrs of a clustr ar clos

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

Business rules FATCA V. 02/11/2015

Business rules FATCA V. 02/11/2015 Elmnt Attribut Siz InputTyp Rquirmnt BUSINESS RULES TYPE ERROR ACK Xpath I.Mssag Hadr FATCA_OECD Vrsion xsd: string = Validation WrongVrsion ftc:fatca_oecd/vrsion SndingCompanyIN Unlimit d xsd: string

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

Sci.Int.(Lahore),26(1),131-138,2014 ISSN 1013-5316; CODEN: SINTE 8 131

Sci.Int.(Lahore),26(1),131-138,2014 ISSN 1013-5316; CODEN: SINTE 8 131 Sci.Int.(Lahor),26(1),131-138,214 ISSN 113-5316; CODEN: SINTE 8 131 REQUIREMENT CHANGE MANAGEMENT IN AGILE OFFSHORE DEVELOPMENT (RCMAOD) 1 Suhail Kazi, 2 Muhammad Salman Bashir, 3 Muhammad Munwar Iqbal,

More information

ImportingCoreInternationalCrimes intonationallaw

ImportingCoreInternationalCrimes intonationallaw ImportngCorIntrnatonalCrms ntonatonallaw MortnBrgsmo,MadsHarlm andnobuohayash(dtors) Importng Cor Intrnatonal Crms nto Natonal Law Mortn Brgsmo, Mads Harlm and Nobuo Hayash (dtors) 2010 Scond Edton Torkl

More information

New Basis Functions. Section 8. Complex Fourier Series

New Basis Functions. Section 8. Complex Fourier Series Nw Basis Functions Sction 8 Complx Fourir Sris Th complx Fourir sris is prsntd first with priod 2, thn with gnral priod. Th connction with th ral-valud Fourir sris is xplaind and formula ar givn for convrting

More information

CPU. Rasterization. Per Vertex Operations & Primitive Assembly. Polynomial Evaluator. Frame Buffer. Per Fragment. Display List.

CPU. Rasterization. Per Vertex Operations & Primitive Assembly. Polynomial Evaluator. Frame Buffer. Per Fragment. Display List. Elmntary Rndring Elmntary rastr algorithms for fast rndring Gomtric Primitivs Lin procssing Polygon procssing Managing OpnGL Stat OpnGL uffrs OpnGL Gomtric Primitivs ll gomtric primitivs ar spcifid by

More information

Rank Optimization of Personalized Search

Rank Optimization of Personalized Search Rank Optmzaton of Personalzed Search Ln LI Zhengl YANG Masar KITSUREGAWA Agmentng the global rankng based on the lnkage strctre of the Web s one of the poplar approaches n data engneerng commnty today

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

ASTIN COLLOQUIUM BERLIN 2003, TOPIC 1: RISK EVALUATION. Exposure Rating in Liability Reinsurance

ASTIN COLLOQUIUM BERLIN 2003, TOPIC 1: RISK EVALUATION. Exposure Rating in Liability Reinsurance ASTIN COLLOQUIUM BERLIN 23, TOPIC : RISK EVALUATION Exposre Ratng n Lablty Rensrance Dr. Thomas Mack and Mchael Fackler Chef Actary Non-Lfe Senor Actary Mnch Rensrance Company Mnch Rensrance Company 879

More information

Electronic Commerce. and. Competitive First-Degree Price Discrimination

Electronic Commerce. and. Competitive First-Degree Price Discrimination Elctronic Commrc and Comptitiv First-Dgr Pric Discrimination David Ulph* and Nir Vulkan ** Fbruary 000 * ESRC Cntr for Economic arning and Social Evolution (ESE), Dpartmnt of Economics, Univrsity Collg

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

The international Internet site of the geoviticulture MCC system Le site Internet international du système CCM géoviticole

The international Internet site of the geoviticulture MCC system Le site Internet international du système CCM géoviticole Th intrnational Intrnt sit of th goviticultur MCC systm L sit Intrnt intrnational du systèm CCM géoviticol Flávio BELLO FIALHO 1 and Jorg TONIETTO 1 1 Rsarchr, Embrapa Uva Vinho, Caixa Postal 130, 95700-000

More information

5.4 Exponential Functions: Differentiation and Integration TOOTLIFTST:

5.4 Exponential Functions: Differentiation and Integration TOOTLIFTST: .4 Eponntial Functions: Diffrntiation an Intgration TOOTLIFTST: Eponntial functions ar of th form f ( ) Ab. W will, in this sction, look at a spcific typ of ponntial function whr th bas, b, is.78.... This

More information

SOFTWARE ENGINEERING AND APPLIED CRYPTOGRAPHY IN CLOUD COMPUTING AND BIG DATA

SOFTWARE ENGINEERING AND APPLIED CRYPTOGRAPHY IN CLOUD COMPUTING AND BIG DATA Intrnational Journal on Tchnical and Physical Problms of Enginring (IJTPE) Publishd by Intrnational Organization of IOTPE ISSN 077-358 IJTPE Journal www.iotp.com ijtp@iotp.com Sptmbr 015 Issu 4 Volum 7

More information

SPECIAL VOWEL SOUNDS

SPECIAL VOWEL SOUNDS SPECIAL VOWEL SOUNDS Plas consult th appropriat supplmnt for th corrsponding computr softwar lsson. Rfr to th 42 Sounds Postr for ach of th Spcial Vowl Sounds. TEACHER INFORMATION: Spcial Vowl Sounds (SVS)

More information

Tax Collection, Transfers, and Corruption: the Russian Federation at the Crossroads 1)

Tax Collection, Transfers, and Corruption: the Russian Federation at the Crossroads 1) ÝÊÎÍÎÌÈ ÅÑÊÈÉ ÆÓÐÍÀË ÂØÝ ¹ 1 2003 3 ÂÎÏÐÎÑÛ ÒÅÎÐÈÈ Tax Collcton, Transfrs, and Corrupton: th Russan Fdraton at th Crossroads 1) M. Mokhtar, I. Grafova Abstract Aftr thortcally consdrng th problms of tax

More information

Section G3: Differential Amplifiers

Section G3: Differential Amplifiers Scton G3: Dffrntal Amplfrs h dffrntal amplfr may b mplmntd usng Js or Fs and s a commonly usd buldng block n analog dsgn. W ar gong to b concntratng on th J mplmntaton of th dffrntal par as mttr-coupld,

More information

Logistic Regression. Lecture 4: More classifiers and classes. Logistic regression. Adaboost. Optimization. Multiple class classification

Logistic Regression. Lecture 4: More classifiers and classes. Logistic regression. Adaboost. Optimization. Multiple class classification Lecture 4: More classfers and classes C4B Machne Learnng Hlary 20 A. Zsserman Logstc regresson Loss functons revsted Adaboost Loss functons revsted Optmzaton Multple class classfcaton Logstc Regresson

More information