Variations on the theme of slacks-based measure of efficiency in DEA

Size: px
Start display at page:

Download "Variations on the theme of slacks-based measure of efficiency in DEA"

Transcription

1 GRIPS Plicy Ifrmati Ceter Dicui Paper : 8-4 Variati the theme f lac-baed meaure f efficiecy i DEA Karu Te Natial Graduate Ititute fr Plicy Studie Rppgi, Miat-u, Ty , Japa te@gripacp Abtract: I DEA, there are typically tw cheme fr meaurig efficiecy f DMU; radial ad -radial Radial mdel aume prprtial chage f iput/utput ad uually remaiig lac are t directly accuted fr iefficiecy O the ther had, -radial mdel deal with lac f each iput/utput idividually ad idepedetly, ad itegrate them it a efficiecy meaure, called lac-baed meaure (SBM) I thi paper, we pit ut hrtcmig f the SBM ad prpe 4 variat f the SBM mdel The rigial SBM mdel evaluate efficiecy f DMU referrig t the furthet frtier pit withi a rage Thi reult i the hardet cre fr the bective DMU ad the precti may g t a remte pit the efficiet frtier which may be iapprpriate a the referece I a effrt t vercme thi hrtcmig, we firt ivetigate frtier (facet) tructure f the prducti pibility et The we prpe Variati I that evaluate each DMU by the earet pit the ame frtier a the SBM fud Hwever, there exit ther ptetial facet fr evaluatig DMU Therefre we prpe Variati II that evaluate each DMU frm all facet We the emply cluterig methd t claify DMU it everal grup, ad apply Variati II withi each cluter Thi Variati III give mre reaable efficiecy cre with le effrt Latly we prpe a radm earch methd (Variati IV) fr reducig the burde f eumerati f facet The reult are apprximate but practical i uage Keywrd: DEA, SBM, facet, eumerati, cluterig, radm earch Reearch upprted by Grat-i-Aid fr Scietific Reearch, Japa Sciety fr the Prmti f Sciece

2 GRIPS Plicy Ifrmati Ceter Dicui Paper : 8-4 Itrducti I mt DEA mdel, the prducti pibility et i a plyhedral cvex et whe vertice crrepd t the efficiet DMU i the mdel A plyhedral cvex et ca be defied by it vertice r by it upprtig hyperplae (Simard [4]) I DEA literature, mai fcu i directed t vertice while cmparatively few reearche are ccered with the upprtig hyperplae Oe f the purpe f thi paper i t fill the gap betwee the tw apprache: vertex ad hyperplae We firtly dicu the characteritic f the upprtig hyperplae t the prducti pibility et i DEA The, baed thi hyperplae, we prpe everal variat f the lac-baed meaure f efficiecy Rughly peaig, we have tw type f meaure i DEA; radial ad -radial Radial meaure are repreeted by CCR [2] ad BCC [] mdel Their drawbac exit i that iput/utput are aumed t uderg prprtial chage ad remaiig lac are t accuted fr i the efficiecy cre N-radial mdel are repreeted by the lac-baed meaure (SBM) [5] The SBM evaluate efficiecy baed the lac-baed meaure t the efficiet frtier Hwever, ice it bective i t miimize thi meaure, the referet pit i apt t be far frm the bective DMU Hwever, there exit ther apprach; t fid the earet pit the frtier Fr thi purpe we firt mdify the SBM t catch the miimum lac-baed meaure pit the facet that the SBM fud fr the DMU We call thi Variati I The, after ivetigati f upprtig hyperplae (facet) t the prducti pibility et, we exted thi apprach t cider all facet, reultig i Variati II Sice the eumerati f facet eed maive cmputati, we prpe tw mre cveiet variati; e cluterig (Variati III) ad the ther radm earch (Variati IV) Thi paper ufld a fllw We itrduce the SBM ad everal prpertie f facet (hyperplae) i Secti 2 The we mdify the SBM i uch a way that itead f miimizati f the bective fucti we maximize it the facet explred by the SBM (Variati I) i Secti 3 We prpe a methd fr fidig all facet f the prducti pibility et i Secti 4 Uig thi reult we exted Variati I t emply all facet (Variati II) i Secti 5 The we implify Variati II t tw cheme; e cluterig (Variati III) ad the ther radm earch (Variati IV) i Secti 6 We mdify ur reult t cpe with the variable retur-t-cale (VRS) evirmet i Secti 7 We cmpare ur variati with the radial (CCR) mdel i Secti 8 Sme ccludig remar fllw i the lat ecti 2 Prelimiarie I thi ecti we itrduce the SBM ad dicu everal prpertie f the facet f prducti pibility et 2 Ntati ad Prducti Pibility Set We deal with DMU (=,,) each havig m iput { xi }( i =, K, m) ad utput { yi}( i =, K, ) We dete the DMU by (, y ) ( =, K, ) ad the iput/utput data matrice by m x X = ( x i ) R ad = ( y i ) R Y, repectively We aume X > ad Y > Uder the ctat retur-t-cale (CRS) aumpti the prducti pibility et i defied by P = {( x, y) x Xλ, y Yλ, λ } () where m λ R i the iteity vectr We itrduce -egative iput ad utput lac R ad R t expre x = Xλ ad y = Yλ (2) 22 Efficiecy ad SBM [Defiiti ] (Efficiet DMU) A DMU ( x, y ) P i called CRS-efficiet if ay luti f the ytem x = Xλ y = Yλ, λ,,, ( x, y ) ha = ad = Otherwie i called CRS-iefficiet, ie there exit -egative but -zer (emi-pitive) lac fr the abve ytem 2

3 GRIPS Plicy Ifrmati Ceter Dicui Paper : 8-4 Thi defiiti crrepd t the Paret-Kpma defiiti f efficiecy: A DMU i fully efficiet if ad ly if it i t pible t imprve ay iput r utput withut wreig me ther iput r utput (Cper et al [3], p 45) The SBM ([5]) lve the fllwig prgram fr DMU ( x, y)( =, K, ) [Theme -- Origial SBM] m i i= mi m xi ρ = mi (3) r r= yr ubect t = = λ ( ) x λ = x y λ = y (4) Thi fractial prgram ca be lved by trafrmig it a equivalet liear prgram (ee [5]) Let a ptimal luti f the SBM be ( λ,, ) [Defiiti 2](Referece et) The referece et fr DMU ( x, y ) i defied by R = { λ >, =, K, } (5) [Defiiti 3](Precti) The precti f DMU ( x, y ) i defied by x = x = x λ R y = y = y λ R (6) [Therem ] The prected DMU ( x, y ) i CRS-efficiet (See Appedix A fr a prf) A the bective fucti (3) ugget, the rigial SBM aim t fid the miimum (the wrt) cre aciated with the relatively maximum lac uder the ctrait (4) Thi might prect the DMU t a very remte pit the frtier (facet) ad metime it i hard t iterpret O the ther had, there i the ppite apprach, ie, t l fr the earet pit the facet, by miimizig the lac-baed meaure frm the frtier Fr thi purpe, we eed t ivetigate the facet f the prducti pibility et, a we hw i the fllwig ecti 23 Facet f Prducti Pibility Set Let ( ξ, η )( =, K, ) be DMU i P We mae a liear cmbiati f thee DMU with pitive 3

4 GRIPS Plicy Ifrmati Ceter Dicui Paper : 8-4 cefficiet a where > ( =, K, ) w ξ η = w ξ = w η L w ξ L w η, (7) [Lemma ] If ay member f ( ξ, η ) ( =, K, ) i CRS-iefficiet, the ξ, ) i CRS-iefficiet ( Prf: Withut lig geerality, we aume that ( ξ, η ) i CRS-iefficiet The, the ytem ξ = Xλ, η = Yλ, λ,, (8) ha a luti ( λ,, ) with (, ) (,) ad (, ) (,) We et By iertig (9) it (7), we have Let u defie Sice ξ = Xλ ad η = Yλ (9) = w w 2 w = ξ ξ ξ = w w 2 w = η η η ξ = wξ w ξ ξ, ) P, ad > ( =, K, ), we have ( η w Hece, we have = 2 η = w η w η = 2 () () ( ξ, η ) P (2) ξ = ξ w η = η w ( ξ, η ) Thu, ha -egative ad -zer lac (, ) agait ( ξ, η ) Hece it i CRS-iefficiet QED A a ctrapiti f Lemma, we have [Therem 2] If ξ, η ) defied by (7) i CRS-efficiet, the ( ξ, η ) ( =, K, ) mut be CRS-efficiet ( We tice that the revere f thi therem i t alway true Nw, we aume we demtrate the fllwig therem ( ξ, η ) i (7) i CRS-efficiet ad [Therem 3] If ( ξ, η ) defied by (7) i CRS-efficiet, the there exit a upprtig hyperplae t P at ( ξ, η ) which al upprt P at ( ξ, η ) ( =, K, ) Prf: By the trg therem f cmplemetarity, there exit dual variable uch that m v R,u R with v >, u > We ca btai uch a trg cmplemetary luti by uig the additive mdel r the -rieted lac-baed meaure (SBM) mdel [3, 5] 4

5 GRIPS Plicy Ifrmati Ceter Dicui Paper : 8-4 Iertig the defiiti f Taig te f w Hece, the hyperplae v ξ v ξ u u η η = v X u Y ( =, K, ) (3) ξ, η ) i (7) it the firt equality i (3), we have w ( v ξ u η ) L ( v ξ u η ) (4) ( w = > ( =, K, ) ad the ecd iequality i (3), the equality (4) hld if ad ly if v ξ u η = ( =, K, ) v x u y = pae thrugh ( ξ, η ) ( =, K, ) ad upprt P QED Thi therem i helpful i idetifyig the facet f P Sice the ytem f equati (5) i hmgeu, if (, u v ) i a luti t (5), the t( v, u ) ( t > ) i al a luti If the ra f the matrix ξ, K, ξ m ( ) R (6) η, K, η i t le tha m, the the cefficiet ( v, u ) i uiquely determied except fr the calar multiplier t, ice the hyperplae v x u y = pae thrugh the rigi ( x=, y = ) ad remaiig m liearly idepedet ( ξ, η ) determie the hyperplae Thi mea that the directi ( v, u ) i the uique rmal t the upprtig hyperplae If the ra f (6) i le tha m, the there exit multiple ( v, u ) fr the ytem (3) [Defiiti 3] (Facet) We call the upprtig hyperplae v x u y defied i Therem 3 a facet f P 3 Variati I Miimizig lac-baed meaure frm the facet The firt variati i a imple mdificati f the baic SBM i the precedig ecti We maximize the bective fucti rather tha miimizati Fr each DMU ( x, y )(, K, ), we lve the SBM mdel i (3-4) If it i iefficiet, we have it referece = et R defied by (5) The prected DMU i efficiet by Therem ad hece the DMU i the referece et are efficiet by Therem 2 Furthermre, by Therem 3, they frm a facet f P We evaluate the miimum lac-baed meaure ad hece the maximum cre the facet a fllw 2 (5) max m ρ = max ubect t m i= r= x y i i r r (7) 2 Simard (966) called uch hyperplae a extremal upprtig ray 5

6 GRIPS Plicy Ifrmati Ceter Dicui Paper : 8-4 R R λ ( ) x λ = x y λ = y (8) Sice we deal with the ame facet a the baic mdel, we have the relatihip: max mi (9) ρ ρ Thi variati demad e additial LP luti fr each iefficiet DMU ad i cmputatially rather eay Hwever, ice the facet defied by R i a itace f facet ad there may be ther facet f P t be cidered i evaluatig the maximum efficiecy f DMU ( x, y ), we eed t w all facet f P We dicu thi ubect i the ext ecti Nw we hw a example f the SBM ad Variati I [Example ] Table exhibit data fr 2 hpital havig tw iput ad tw utput Iput: Number f dctr ad ure Output: Number f utpatiet ad ipatiet << Iert Table here Table : Data f 2 hpital>> Firt, we lved thi cae by the SBM i (3-4) The, wig the referece et ad hece a facet f iefficiet DMU, we lved the Variati I i (7-8) The reult are diplayed i Table 2 Every iefficiet DMU imprved mi their efficiecy except H Fr example, Hpital C i iefficiet 8265 by the SBM ad it referece are B ρ = ad L We lved the maximum prblem (Variati I) the facet paed by B ad L, ad btaied with the referece B The differece i the gap betwee the max ad the mi bective value meaured by (7) << Iert Table 2 here Table 2: Reult f SBM ad Variati I>> 4 Eumerati f facet I thi ecti, we prpe a methd fr eumeratig all facet f P Let P = ( ξ, η )( =, K, K) be the CRS-efficiet DMU i P C ρ = 855 [Defiiti 3] (fried) A ubet P, K, } f { P } = {( ξ, η )} ( =, K, K) i called fried if a liear cmbiati with pitive cefficiet f { P { P, K, P } i CRS-efficiet max C [Defiiti 4] (maximal fried) A fried i called maximal if ay additi f P (t i the fried) t the fried i mre fried [Defiiti 5] (dmiated fried) A fried i dmiated by ther fried if the et f DMU i a ubet f ther 6

7 GRIPS Plicy Ifrmati Ceter Dicui Paper : 8-4 We prpe a algrithm fr fidig the maximal fried f P = ( ξ, η )( =, K, K) [Algrithm A] Begi Fr = t K Fid_Maximal_Fried f P Next Delete dmiated fried frm the et f fried Obtai the et f facet frm the fial et f fried Ed Subrutie Fid_Maximal_Fried f P Exclude P, K, P frm the cadidate f fried Eumerate all fried f P Remve dmiated fried frm the et f fried Exit ub Let the umber f facet thu geerated be H We have H facet t P: ( h) ( h) Facet( h) : v x u y ( h =, K, H) (2) Facet(h) pae thrugh it fried ad upprt P The abve facet cit f geuie efficiet frtier f the prducti pibility et P Hwever, P ha -efficiet budarie a we ee i Figure a example I the figure lie egmet AB ad BC are efficiet facet, while AD ad CE are -efficiet budarie f P WE tice that, i thi paper, we berve ad deal ly with the efficiet prti f the budary << Iert Figure Figure : Efficiet ad -efficiet frtier>> [Therem 4] Fr every efficiet frtier pit f P, there exit a Facet(h) that tuche the efficiet pit Prf: Every efficiet frtier pit ca be expreed by a pitive liear cmbiati f a et f efficiet vertice f P By ctructi f the maximal_fried i the Algrithm A, the et a well a the efficiet pit i me Facet(h) QED 5 Variati II Miimizig the SBM frm all facet We deal with a et f DMU defied i Secti 2 Step Fidig Efficiet DMU Slve the -rieted SBM mdel r the additive mdel ad fid the et f efficiet DMU Let the et be {( ξ, η ) =, K, K )} where K i the umber f efficiet DMU Step 2 Eumerati f Facet Eumerate all facet applyig Algrithm A i Secti 4 Let the umber f facet thu btaied be H We deal with ly facet i the maximal fried Step 3 Evaluati f Iefficiet DMU Fr a iefficiet DMU x, y ) we evaluate it efficiecy cre a fllw ( 7

8 GRIPS Plicy Ifrmati Ceter Dicui Paper : 8-4 Fr each Facet( h) ( h =, K, H), we lve the fllwig fractial prgram: ρ ( h) = ubect t m max R ( h) R ( h) λ ( ), ξ λ η λ where R(h) i the et f efficiet DMU that pa Facet(h) We btai the efficiecy cre f DMU x, y ) a ( m i= r= ( h) { } x y i i r r = x = y (2) (22) all ρ = max ρ (23) h We have the fllwig iequalitie amg the three cre: all max mi ρ ρ ρ (24) [Example 2] I the abve example, the et f fried cmped f tw DMU are fud t be AD, BD, AL, BL ad DL The et f fried cmped f three DMU are ADL ad BDL The et ABDL cat be a fried (facet) Hece the maximal fried are ADL ad BDL Uig ADL ad BDL a referece repectively, we lved the prgram (2-22), ad btaied the efficiecy cre fr iefficiet DMU a exhibited i Table 3 Fr example, fr DMU E, we have BDL ρ = 76823(with referece A) ad ρ = 75236(with referece D, L) Thu ρ all = ADL E E E <<Iert Table 3 here Table 3: Reult f SBM ad Variati II>> Cmpari with Table 2 reveal everal iteretig feature f Variati II A demtrated i (24), the efficiecy cre f Variati II i t le tha the f the SBM ad Variati I fr each DMU 6 Hw t reduce a maive eumerati I Variati II, the eumerati f facet eed a ermu cmputati time ad pace fr large cale prblem, eve thugh advace i recet IT techlgie are amazig i bth apect If we have m=6 (# f iput), =5 (# f utput) ad =2 (# f efficiet DMU), the i the wrt cae we might eumerate abut 2 C =84,756 cae Of cure, mt f them wuld be fud t be a iefficiet cmbiati I thi ecti we prpe tw mdified veri f Variati II which are le time ad pace cumig 6 Variati III Cluterig Step Cluterig DMU Uig me cluterig methd, we claify all DMU i cluter, ay, Cluter t Cluter L Step 2 Fidig efficiet DMU Thi tep i the ame a the Step f the Variati II Step 3 Evaluatig efficiecy cre fr a iefficiet DMU If the iefficiet DMU x, y ) belg t Cluter h, pic up the efficiet DMU i Cluter h If e f DMU i ( Cluter h i efficiet, we pic up the efficiet DMU i the adacet cluter 8

9 GRIPS Plicy Ifrmati Ceter Dicui Paper : 8-4 Let the ubet f efficiet DMU crrepdig t Cluter h be E( h) = {( ξ, η ), K, ( ξ J, η J )} We create the facet cmped f the efficiet DMU i E(h) uig the ame prcedure a decribed i Step 2 f the precedig ecti We evaluate the efficiecy f DMU x, y ) i referece t the facet thu btaied i the ( ame way a the Step 3 f the precedig ecti If the prgram (2-22) ha feaible luti, DMU ( x, y ) i udged t be efficiet i thi cluter, ie, it i glbally iefficiet but lcally efficiet The merit f thi mdificati are a fllw: () By itrducig a ciderable umber f cluter, we ca reduce the umber f the cadidate cmbiati (2) Fr iefficiet DMU, the efficiecy cre i btaied i referece t the efficiet DMU i the ame cluter Thu, the reult are mre acceptable ad udertadable [Example 3] We claified 2 hpital i Table it tw cluter depedig their ize (umber f dctr ad ipatiet) a decribed i the clum Cluter f Table 4 We lved -rieted SBM mdel ad fud 4 efficiet DMU (A,, B, D, L) ad 8 iefficiet DMU with their referece a exhibited i the left ide f Table 4 where we fud everal iapprpriate referece Fr example, C ha referece B ad L, wherea L i t i the ame cluter a C I the cluter, the maximal fried are AD ad BD, ad i the cluter 2 we have ly e facet L Fially, we lved the efficiecy f iefficiet DMU referrig t the facet i the ame cluter ad fud the reult recrded i the right half f Table 4 DMU C ha it referece D ad efficiecy cre which wa upgraded frm the SBM cre 826 DMU i the cluter 2 were all evaluated their efficiecy agait L We fud ifeaibility fr G ad J Hece, we udged them efficiet i thi cluter They are glbally iefficiet but lcally efficiet << Iert Table 4 here Table 4: SBM ad Cluterig reult (Variati III)>> 62 Variati IV Radm Search I thi ecti we prpe a apprximate methd fr fidig facet Step Fidig ceter f gravity f efficiet DMU Let the et f efficiet DMU be P = ( ξ, η )( =, K, K) We calculate their ceter f gravity G a x = ( ξ L ξ )/ K G y = ( η L η )/ K G Figure 2 illutrate a example We te that we ca utilize ay pitive liear cmbiati f efficiet DMU itead f the ceter f gravity fr ur purpe Step 2 Creatig radm directi arud efficiet DMU Fr each efficiet DMU P = ( ξ, η ) we cmpute the directi frm G t P = ( ξ, η ) a ( ξ x, η y ) ad the perturb the directi lightly uig radm umber Let the directi thu perturbed be G G ( d, d ) x y Step 3 Fidig a facet We lve the fllwig liear prgram i t Rad K λ R : K K (25) 9

10 GRIPS Plicy Ifrmati Ceter Dicui Paper : 8-4 max t ubec t x d t ξ λ L ξ λ G x K K y d t ηλ L η λ G y K t, λ K (26) Let a ptimal luti be ( t, λ ) If t =, the the ceter G i efficiet ad all P = ( ξ, η)( =, K, K) are fried Thi cae ha ly e efficiet facet by Therem 3 If t >, the the referece DMU crrepdig t pitive λ frm a facet f P, ice the ptimal luti i btaied a budary f P Step 4 Repeatig the radm earch We repeat the radm earch arud the K efficiet DMU util a ufficiet umber f facet i fud Step 5 Evaluatig iefficiet DMU We evaluate the efficiecy cre f iefficiet DMU uig the facet thu fud i the ame maer a the Variati II <<Iert Figure 2 here Figure 2: Radm earch arud efficiet DMU>> [Example 4] I the hpital example, DMU A, B, D ad L are efficiet Table 5 dete their ceter f gravity ad directi vectr frm the ceter t A, B, D ad L We diturb thee vectr radmly ad, fr example fr D, we have, dx=7, dx2=-3, dy=8, dy2=-3 Uig thi directi we lved the prgram (26) ad btaied λ = 3822, λ = 455, λ = Thu, ADL pa a facet f P I thi way we ca fid A D L facet f P apprximately Table 6 exhibit reult f radm earche We tried tw radm earche (perturbed directi) fr each efficiet DMU a diplayed i the table Evetually we fud the tw maximal fried (facet); ADL ad BDL The rea why we perturb the directi arud vertice i that everal facet are cected at a vertex ad we ca fid facet with high prbability <<Iert Table 5 here Table 5: Ceter ad directi>> Iert Table 6 here Table 6: Reult f radm earch>> 7 Variable retur-t-cale (VRS) cae S far we have dicued the ctat retur t cale cae We eed me alterati i the variable retur-t-cale (VRS) cae, which require the cvexity cditi the iteity vectr λ R : L = (27) λ λ I thi ecti, we preet ly imprtat addeda t the precedig ecti The prducti pibility et () ad the SBM mdel (4) have the additial ctrait (27) 2 Equati (7) i mdified t:

11 GRIPS Plicy Ifrmati Ceter Dicui Paper : Lemma tur ut t: ξ = wξ L w ξ η = wη L w η w L w =, w > ( ) (7A) [Lemma A] If ay member f ( ξ, η ) ( =, K, ) i VRS-iefficiet, the ξ, ) i VRS-iefficiet ( Prf: Withut lig geerality, we aume that ( ξ, η ) i VRS-iefficiet The, the ytem ξ = Xλ, η = Yλ eλ =, λ,, ha a luti ( λ,, ) with (, ) (,) ad (, ) (,), where e i the rw vectr with all elemet equal t We et ξ = Xλ ad η = Yλ (9A) By iertig (9) it (7), we have Let u defie Sice = w w 2 w = ξ ξ ξ = w w 2 w = η η η ξ = wξ w ξ = 2 η = w η w η = 2 ξ, ) P ad w =, w > ( =, K, ), we have ( η Hece, we have = (A) (A) ( ξ, η ) P (2A) ξ = ξ w η = η w ( ξ, η ) Thu, ha -egative ad -zer lac (, ) agait ( ξ, η ) Hece it i VRS-iefficiet QED 4 Therem 3 chage t: [Therem 3A] If ( ξ, η ) defied by (7A) i VRS-efficiet, the there exit a upprtig hyperplae t P at ( ξ, η ) which al upprt P at ( ξ, η ) ( =, K, ) Prf: By the trg therem f cmplemetarity, there exit dual variable v >, u > uch that v ξ v ξ u u η η u = v X u Y u e u m v R,u R, u R with ( =, K, ) (3A)

12 GRIPS Plicy Ifrmati Ceter Dicui Paper : 8-4 Iertig the defiiti f ξ, η ) i (7A) it the firt equality i (3A) ad tig =, we have Taig te f w ( w = w ( v ξ u η u ) L w ( v ξ u η u ) = (4A) > ( =, K, ) ad the ecd iequality i (3A), the equality (4A) hld if ad ly if (5A) v ξ u η = ( =, K, ) Hece the hyperplae v x u y u = pae thrugh ( ξ, η ) ( =, K, ) ad upprt P QED Sice the ytem f equati (5A) i hmgeu, if t( v, u, u ) ( t > ) i al a luti If the ra f the matrix ξ, K, ξ R η, K, η i t le tha m, the the cefficiet ( v, u, u ( m ) ( v, u, u ) i a luti t (5A), the (6A) ) i uiquely determied except fr the calar multiplier t Thi mea that the directi ( v, u ) i the uique rmal t the upprtig hyperplae If the ra f (6A) i le tha m, the there exit multiple ( v, u, u ) fr the ytem (3A) Defiiti (Facet) We call the upprtig hyperplae v x u y u defied i Therem A a facet f P I what fllw, we che the ceter f gravity f ( ξ, η )( =, K, ) a ( ξ, η ), ie w / ( ) = 5 We add the cvexity cditi eλ =t the liear prgram (26) 8 Cmpari with the radial mdel We cmpared the cre btaied by the SBM, Variati II ad the radial CCR mdel a diplayed i Table 7 The CCR cre i t le tha that f the SBM ([3, p ]) Hwever, Variati II ad the CCR are mixed We have theretical evidece betwee the tw The reult idicate vlatility f cre ad ra depedig the mdel, ad cte the imprtace f mdel electi a i alway the cae i DEA applicati 9 Ccludig remar <<Iert Table 7 here Table 7: Cmpari f SBM, Variati II ad CCR>> I thi paper, we have prped 4 variat f the SBM They have cmm characteritic a fllw: They are uit-ivariat, ie the cre are idepedet f the uit i which the iput ad utput are meaured prvided thee uit are the ame fr every DMU 2 We ca impe weight exgeuly t each iput/utput depedig their imprtace, eg ct hare Refer t Cper et al [4, p5] ad Tutui ad Gt [7] 3 Althugh we have develped ur mdel i the -called -rieted veri, ie bth iput ad utput iefficiecie are accuted i the efficiecy evaluati, we ca deal with the iput (utput) rieted mdel by taig the umeratr (demiatr) f the bective fucti (3, 7, 2) a the target 4 Future reearch ubect iclude (a) experimet real-wrld large cale prblem ad (b) extei t the uper-sbm mdel [6] Referece [] Baer RD, Chare A, Cper WW (984) Sme methd fr etimatig techical ad cale efficiecie i DEA, Maagemet Sciece 984; 3:

13 GRIPS Plicy Ifrmati Ceter Dicui Paper : 8-4 [2] Chare A, Cper WW, Rhde E (978) Meaurig the efficiecy f decii maig uit Eurpea Jural f Operatial Reearch, 2, [3] Cper WW, Seifrd LM, Te K (27) Data evelpmet aalyi: A cmpreheive text with mdel, applicati, referece ad DEA-Slver ftware, 2d Editi, Spriger [4] Simard M, (966) Liear prgrammig, tralated by Jewell WS, Pretice-Hall [5] Te K (2) A lac-baed meaure f efficiecy i data evelpmet aalyi, Eurpea Jural f Operatial Reearch, 3, [6] Te K (22) A lac-baed meaure f uper-efficiecy i data evelpmet aalyi, Eurpea Jural f Operatial Reearch, 43, 32-4 [7] Tutui M, Gt M (28) A multi-divii efficiecy evaluati f US electric pwer cmpaie uig a weighted lac-baed meaure, Sci Ecmic Plaig Sciece, i pre Appedix A Prf f Therem Suppe that ( x, y ) i CRS-iefficiet The there exit a ptimal luti ad -egative lac (, ) fr the prgram: (, λ,, ) with -zer ρ Iertig (6) t (4B) we have: ρ = mi m ubect t = = λ ( ) = = m i= r= x λ = x y λ = y x y i i r r = x λ x = y λ y Fr thi maipulati, we have the bective fucti value fr ( x, y ), Sice (, ) i emi-pitive, we have: m ρ = ρ < ρ mi m i i i= xi r r r= y r (3B) (4B) (5B) (6B) 3

14 GRIPS Plicy Ifrmati Ceter Dicui Paper : 8-4 Thi ctradict the ptimality f ρ QED mi 4

15 GRIPS Plicy Ifrmati Ceter Dicui Paper : 8-4 Table : Data f 2 hpital Iput Output DMU Dctr Nure Outpatiet Ipatiet A B C D E F G H I J K L Table 2: Reult f SBM ad Variati I DMU SBM Ref Variati I Ref A A A B B B C B,L B D D D E B,L L F A,L L G B,L B,L H L L I A,L L J B,L B K B,L L L L L 5

16 GRIPS Plicy Ifrmati Ceter Dicui Paper : 8-4 Table 3: Reult f SBM ad Variati II DMU SBM Ref Variati II Ref A A A B B B C B,L D D D D E B,L A F A,L D G B,L D H L 898 D I A,L A,D,L J B,L D K B,L A,D L L L Table 4: SBM ad Cluterig reult (Variati III) DMU SBM Ref Cluter Variati III Ref Remar A A A B B B C 826 B,L D D D D E 728 B,L A F 686 A,L L G 877 B,L 2 G lcally eff H 77 L 898 D I 92 A,L L J 765 B,L 2 J lcally eff K 862 B,L L L L 2 L 6

17 GRIPS Plicy Ifrmati Ceter Dicui Paper : 8-4 Table 5: Ceter ad directi DMU (I)Dctr (I)Nure (O)Outpatiet (O)Ipatiet A B D L Ceter directi dx dx2 dy dy2 A B D L Table 6: Reult f radm earch DMU dx dx2 dy dy2 Facet fud A A A AL B BL B BDL D ADL D BD L BL L ADL Table 7: Cmpari f SBM, Variati II ad CCR DMU SBM Ref Ra Variati II Ref Ra CCR Ref Ra A A A A B B B B C B,L D B,D 8 D D D D E B,L A A,D,L 2 F A,L D B,D G B,L D B,L 7 H L D A,D,L I A,L A,D,L B,L 5 J B,L D D 9 K B,L A,D A,D 6 L L L L 7

18 GRIPS Plicy Ifrmati Ceter Dicui Paper : 8-4 Iput 2 D A B C E Iput Figure : Efficiet ad -efficiet frtier Iput 2 A Ceter G B C Iput Figure 2: Radm earch arud efficiet DMU 8

An epsilon-based measure of efficiency in DEA revisited -A third pole of technical efficiency-

An epsilon-based measure of efficiency in DEA revisited -A third pole of technical efficiency- GRIPS Plicy Ifrmati Ceter Discussi Paper : 09-2 A epsil-based measure f efficiecy i DEA revisited -A third ple f techical efficiecy- Karu Te Natial Graduate Istitute fr Plicy Studies 7-22- Rppgi, Miat-ku,

More information

Multi-objective Programming Approach for. Fuzzy Linear Programming Problems

Multi-objective Programming Approach for. Fuzzy Linear Programming Problems Applied Mathematical Scieces Vl. 7 03. 37 8-87 HIKARI Ltd www.m-hikari.cm Multi-bective Prgrammig Apprach fr Fuzzy Liear Prgrammig Prblems P. Padia Departmet f Mathematics Schl f Advaced Scieces VIT Uiversity

More information

On the Consistency of Slacks-based Measure-Max Model and Super-Slacks-based Measure Model

On the Consistency of Slacks-based Measure-Max Model and Super-Slacks-based Measure Model GRIPS Dicui Paper 6-24 O the Citecy Slack-baed Meaure-Max Mdel ad Super-Slack-baed Meaure Mdel Karu Te Nveber 206 Natial Graduate Ititute r Plicy Studie 7-22- Rppgi, Miat-ku, Tky, Japa 06-8677 O the Citecy

More information

Super-efficiency Models, Part II

Super-efficiency Models, Part II Super-efficiec Mdels, Part II Emilia Niskae The 4th f Nvember S steemiaalsi Ctets. Etesis t Variable Returs-t-Scale (0.4) S steemiaalsi Radial Super-efficiec Case Prblems with Radial Super-efficiec Case

More information

Fourier Method for Solving Transportation. Problems with Mixed Constraints

Fourier Method for Solving Transportation. Problems with Mixed Constraints It. J. Ctemp. Math. Scieces, Vl. 5, 200,. 28, 385-395 Furier Methd fr Slvig Trasprtati Prblems with Mixed Cstraits P. Padia ad G. Nataraja Departmet f Mathematics, Schl f Advaced Scieces V I T Uiversity,

More information

Chapter 5. Root Locus Techniques

Chapter 5. Root Locus Techniques Chapter 5 Rt Lcu Techique Itrducti Sytem perfrmace ad tability dt determied dby cled-lp l ple Typical cled-lp feedback ctrl ytem G Ope-lp TF KG H Zer -, - Ple 0, -, - K Lcati f ple eaily fud Variati f

More information

Strategy in practice: a quantitative approach to target setting

Strategy in practice: a quantitative approach to target setting MPRA Muich Peral RePEc Archive Strategy i practice: a quatitative apprach t target ettig Iree Fafaliu ad Paagiti Zervpul Uiverity f Piraeu, Ope Uiverity f Cypru 4. Jauary 2014 Olie at http://mpra.ub.ui-mueche.de/54054/

More information

STRUCTURES IN MIKE 21. Flow over sluice gates A-1

STRUCTURES IN MIKE 21. Flow over sluice gates A-1 A-1 STRUCTURES IN MIKE 1 Fl ver luice gate Fr a give gemetry f the luice gate ad k ater level uptream ad dtream f the tructure, the fl rate, ca be determied thrugh the equati f eergy ad mmetum - ee B Pedere,

More information

Portfolio Performance Evaluation in a Modified Mean-Variance-Skewness Framework with Negative Data

Portfolio Performance Evaluation in a Modified Mean-Variance-Skewness Framework with Negative Data Available lie at http://idea.srbiau.ac.ir It. J. Data Evelpmet Aalysis (ISSN 345-458X) Vl., N.3, Year 04 Article ID IJDEA-003,3 pages Research Article Iteratial Jural f Data Evelpmet Aalysis Sciece ad

More information

Chapter 3.1: Polynomial Functions

Chapter 3.1: Polynomial Functions Ntes 3.1: Ply Fucs Chapter 3.1: Plymial Fuctis I Algebra I ad Algebra II, yu ecutered sme very famus plymial fuctis. I this secti, yu will meet may ther members f the plymial family, what sets them apart

More information

A New Method for Finding an Optimal Solution. of Fully Interval Integer Transportation Problems

A New Method for Finding an Optimal Solution. of Fully Interval Integer Transportation Problems Applied Matheatical Scieces, Vl. 4, 200,. 37, 89-830 A New Methd fr Fidig a Optial Sluti f Fully Iterval Iteger Trasprtati Prbles P. Padia ad G. Nataraja Departet f Matheatics, Schl f Advaced Scieces,

More information

ENGI 4421 Central Limit Theorem Page Central Limit Theorem [Navidi, section 4.11; Devore sections ]

ENGI 4421 Central Limit Theorem Page Central Limit Theorem [Navidi, section 4.11; Devore sections ] ENGI 441 Cetral Limit Therem Page 11-01 Cetral Limit Therem [Navidi, secti 4.11; Devre sectis 5.3-5.4] If X i is t rmally distributed, but E X i, V X i ad is large (apprximately 30 r mre), the, t a gd

More information

are specified , are linearly independent Otherwise, they are linearly dependent, and one is expressed by a linear combination of the others

are specified , are linearly independent Otherwise, they are linearly dependent, and one is expressed by a linear combination of the others Chater 3. Higher Order Liear ODEs Kreyszig by YHLee;4; 3-3. Hmgeeus Liear ODEs The stadard frm f the th rder liear ODE ( ) ( ) = : hmgeeus if r( ) = y y y y r Hmgeeus Liear ODE: Suersiti Pricile, Geeral

More information

STRONG DEVIATION THEOREMS FOR THE SEQUENCE OF CONTINUOUS RANDOM VARIABLES AND THE APPROACH OF LAPLACE TRANSFORM

STRONG DEVIATION THEOREMS FOR THE SEQUENCE OF CONTINUOUS RANDOM VARIABLES AND THE APPROACH OF LAPLACE TRANSFORM Joural of Statitic: Advace i Theory ad Applicatio Volume, Number, 9, Page 35-47 STRONG DEVIATION THEORES FOR THE SEQUENCE OF CONTINUOUS RANDO VARIABLES AND THE APPROACH OF LAPLACE TRANSFOR School of athematic

More information

MATH Midterm Examination Victor Matveev October 26, 2016

MATH Midterm Examination Victor Matveev October 26, 2016 MATH 33- Midterm Examiati Victr Matveev Octber 6, 6. (5pts, mi) Suppse f(x) equals si x the iterval < x < (=), ad is a eve peridic extesi f this fucti t the rest f the real lie. Fid the csie series fr

More information

The Excel FFT Function v1.1 P. T. Debevec February 12, The discrete Fourier transform may be used to identify periodic structures in time ht.

The Excel FFT Function v1.1 P. T. Debevec February 12, The discrete Fourier transform may be used to identify periodic structures in time ht. The Excel FFT Fucti v P T Debevec February 2, 26 The discrete Furier trasfrm may be used t idetify peridic structures i time ht series data Suppse that a physical prcess is represeted by the fucti f time,

More information

Imprecise DEA for Setting Scale Efficient Targets

Imprecise DEA for Setting Scale Efficient Targets Int. Jurnal f Math. Anali, Vl. 3, 2009, n. 6, 747-756 Imprecie DEA fr Setting Scale Efficient Target N. Malekmhammadi a, F. Heinzadeh Ltfi b and Azmi B Jaafar c a Intitute fr Mathematical Reearch, Univeriti

More information

Axial Temperature Distribution in W-Tailored Optical Fibers

Axial Temperature Distribution in W-Tailored Optical Fibers Axial Temperature Distributi i W-Tailred Optical ibers Mhamed I. Shehata (m.ismail34@yah.cm), Mustafa H. Aly(drmsaly@gmail.cm) OSA Member, ad M. B. Saleh (Basheer@aast.edu) Arab Academy fr Sciece, Techlgy

More information

Fig. 1: Streamline coordinates

Fig. 1: Streamline coordinates 1 Equatio of Motio i Streamlie Coordiate Ai A. Soi, MIT 2.25 Advaced Fluid Mechaic Euler equatio expree the relatiohip betwee the velocity ad the preure field i ivicid flow. Writte i term of treamlie coordiate,

More information

Quantum Mechanics for Scientists and Engineers. David Miller

Quantum Mechanics for Scientists and Engineers. David Miller Quatum Mechaics fr Scietists ad Egieers David Miller Time-depedet perturbati thery Time-depedet perturbati thery Time-depedet perturbati basics Time-depedet perturbati thery Fr time-depedet prblems csider

More information

IJISET - International Journal of Innovative Science, Engineering & Technology, Vol. 2 Issue 12, December

IJISET - International Journal of Innovative Science, Engineering & Technology, Vol. 2 Issue 12, December IJISET - Iteratial Jural f Ivative Sciece, Egieerig & Techlgy, Vl Issue, December 5 wwwijisetcm ISSN 48 7968 Psirmal ad * Pararmal mpsiti Operatrs the Fc Space Abstract Dr N Sivamai Departmet f athematics,

More information

Société de Calcul Mathématique, S. A. Algorithmes et Optimisation

Société de Calcul Mathématique, S. A. Algorithmes et Optimisation Société de Calcul Mathématique S A Algorithme et Optimiatio Radom amplig of proportio Berard Beauzamy Jue 2008 From time to time we fid a problem i which we do ot deal with value but with proportio For

More information

Directional Duality Theory

Directional Duality Theory Suther Illiis Uiversity Carbdale OpeSIUC Discussi Papers Departmet f Ecmics 2004 Directial Duality Thery Daiel Primt Suther Illiis Uiversity Carbdale Rlf Fare Oreg State Uiversity Fllw this ad additial

More information

Intermediate Division Solutions

Intermediate Division Solutions Itermediate Divisi Slutis 1. Cmpute the largest 4-digit umber f the frm ABBA which is exactly divisible by 7. Sluti ABBA 1000A + 100B +10B+A 1001A + 110B 1001 is divisible by 7 (1001 7 143), s 1001A is

More information

A Study on Estimation of Lifetime Distribution with Covariates Under Misspecification

A Study on Estimation of Lifetime Distribution with Covariates Under Misspecification Prceedigs f the Wrld Cgress Egieerig ad Cmputer Sciece 2015 Vl II, Octber 21-23, 2015, Sa Fracisc, USA A Study Estimati f Lifetime Distributi with Cvariates Uder Misspecificati Masahir Ykyama, Member,

More information

Markov processes and the Kolmogorov equations

Markov processes and the Kolmogorov equations Chapter 6 Markv prcesses ad the Klmgrv equatis 6. Stchastic Differetial Equatis Csider the stchastic differetial equati: dx(t) =a(t X(t)) dt + (t X(t)) db(t): (SDE) Here a(t x) ad (t x) are give fuctis,

More information

Ch. 1 Introduction to Estimation 1/15

Ch. 1 Introduction to Estimation 1/15 Ch. Itrducti t stimati /5 ample stimati Prblem: DSB R S f M f s f f f ; f, φ m tcsπf t + φ t f lectrics dds ise wt usually white BPF & mp t s t + w t st. lg. f & φ X udi mp cs π f + φ t Oscillatr w/ f

More information

Grade 3 Mathematics Course Syllabus Prince George s County Public Schools

Grade 3 Mathematics Course Syllabus Prince George s County Public Schools Ctet Grade 3 Mathematics Curse Syllabus Price Gerge s Cuty Public Schls Prerequisites: Ne Curse Descripti: I Grade 3, istructial time shuld fcus fur critical areas: (1) develpig uderstadig f multiplicati

More information

u t u 0 ( 7) Intuitively, the maximum principles can be explained by the following observation. Recall

u t u 0 ( 7) Intuitively, the maximum principles can be explained by the following observation. Recall Oct. Heat Equatio M aximum priciple I thi lecture we will dicu the maximum priciple ad uiquee of olutio for the heat equatio.. Maximum priciple. The heat equatio alo ejoy maximum priciple a the Laplace

More information

Isolated Word Recogniser

Isolated Word Recogniser Lecture 5 Iolated Word Recogitio Hidde Markov Model of peech State traitio ad aligmet probabilitie Searchig all poible aligmet Dyamic Programmig Viterbi Aligmet Iolated Word Recogitio 8. Iolated Word Recogier

More information

BIO752: Advanced Methods in Biostatistics, II TERM 2, 2010 T. A. Louis. BIO 752: MIDTERM EXAMINATION: ANSWERS 30 November 2010

BIO752: Advanced Methods in Biostatistics, II TERM 2, 2010 T. A. Louis. BIO 752: MIDTERM EXAMINATION: ANSWERS 30 November 2010 BIO752: Advaced Methds i Bistatistics, II TERM 2, 2010 T. A. Luis BIO 752: MIDTERM EXAMINATION: ANSWERS 30 Nvember 2010 Questi #1 (15 pits): Let X ad Y be radm variables with a jit distributi ad assume

More information

ENGI 4421 Central Limit Theorem Page Central Limit Theorem [Navidi, section 4.11; Devore sections ]

ENGI 4421 Central Limit Theorem Page Central Limit Theorem [Navidi, section 4.11; Devore sections ] ENGI 441 Cetral Limit Therem Page 11-01 Cetral Limit Therem [Navidi, secti 4.11; Devre sectis 5.3-5.4] If X i is t rmally distributed, but E X i, V X i ad is large (apprximately 30 r mre), the, t a gd

More information

If σis unknown. Properties of t distribution. 6.3 One and Two Sample Inferences for Means. What is the correct multiplier? t

If σis unknown. Properties of t distribution. 6.3 One and Two Sample Inferences for Means. What is the correct multiplier? t /8/009 6.3 Oe a Tw Samle Iferece fr Mea If i kw a 95% Cfiece Iterval i 96 ±.96 96.96 ± But i ever kw. If i ukw Etimate by amle taar eviati The etimate taar errr f the mea will be / Uig the etimate taar

More information

Fourier Series & Fourier Transforms

Fourier Series & Fourier Transforms Experimet 1 Furier Series & Furier Trasfrms MATLAB Simulati Objectives Furier aalysis plays a imprtat rle i cmmuicati thery. The mai bjectives f this experimet are: 1) T gai a gd uderstadig ad practice

More information

Alkaline Surfactant Polymer alternating with Miscible CO 2 (ASPaM) Seyed Hamidreza Ghazizadeh Behzadi And Dr. Brian F. Towler

Alkaline Surfactant Polymer alternating with Miscible CO 2 (ASPaM) Seyed Hamidreza Ghazizadeh Behzadi And Dr. Brian F. Towler Alkalie urfactat Plymer alteratig with Micible CO 2 (APaM) eyed Hamidreza Ghazizadeh Behzadi Ad Dr. Bria F. Twler OUTLINE Why AP ad why WAG APaM advatage Hw t imulate the APaM ectr Mdel Reult Cclui Why

More information

A Hartree-Fock Calculation of the Water Molecule

A Hartree-Fock Calculation of the Water Molecule Chemistry 460 Fall 2017 Dr. Jea M. Stadard Nvember 29, 2017 A Hartree-Fck Calculati f the Water Mlecule Itrducti A example Hartree-Fck calculati f the water mlecule will be preseted. I this case, the water

More information

20. CONFIDENCE INTERVALS FOR THE MEAN, UNKNOWN VARIANCE

20. CONFIDENCE INTERVALS FOR THE MEAN, UNKNOWN VARIANCE 20. CONFIDENCE INTERVALS FOR THE MEAN, UNKNOWN VARIANCE If the populatio tadard deviatio σ i ukow, a it uually will be i practice, we will have to etimate it by the ample tadard deviatio. Sice σ i ukow,

More information

A Multilevel Cartesian Non-uniform Grid Time Domain Algorithm

A Multilevel Cartesian Non-uniform Grid Time Domain Algorithm A Multilevel Carteia N-uifrm Grid Time Dmai Algrithm Ju Meg 1, Amir ag, Vitaliy Lmaki 3*, ad Eric Michiele 4 1 Departmet f Electrical ad Cmputer Egieerig, Uiverity f Illii at Urbaa Champaig, Urbaa, IL

More information

EECE 301 Signals & Systems Prof. Mark Fowler

EECE 301 Signals & Systems Prof. Mark Fowler EECE 30 Sigal & Sytem Prof. Mark Fowler Note Set #8 C-T Sytem: Laplace Traform Solvig Differetial Equatio Readig Aigmet: Sectio 6.4 of Kame ad Heck / Coure Flow Diagram The arrow here how coceptual flow

More information

Solutions. Definitions pertaining to solutions

Solutions. Definitions pertaining to solutions Slutis Defiitis pertaiig t slutis Slute is the substace that is disslved. It is usually preset i the smaller amut. Slvet is the substace that des the disslvig. It is usually preset i the larger amut. Slubility

More information

LECTURE 13 SIMULTANEOUS EQUATIONS

LECTURE 13 SIMULTANEOUS EQUATIONS NOVEMBER 5, 26 Demad-upply ytem LETURE 3 SIMULTNEOUS EQUTIONS I thi lecture, we dicu edogeeity problem that arie due to imultaeity, i.e. the left-had ide variable ad ome of the right-had ide variable are

More information

Applied Mathematical Sciences, Vol. 9, 2015, no. 3, HIKARI Ltd,

Applied Mathematical Sciences, Vol. 9, 2015, no. 3, HIKARI Ltd, Applied Mathematical Sciece Vol 9 5 o 3 7 - HIKARI Ltd wwwm-hiaricom http://dxdoiorg/988/am54884 O Poitive Defiite Solutio of the Noliear Matrix Equatio * A A I Saa'a A Zarea* Mathematical Sciece Departmet

More information

ELEC 372 LECTURE NOTES, WEEK 4 Dr. Amir G. Aghdam Concordia University

ELEC 372 LECTURE NOTES, WEEK 4 Dr. Amir G. Aghdam Concordia University ELEC 37 LECTURE NOTES, WEE 4 Dr Amir G Aghdam Cocordia Uiverity Part of thee ote are adapted from the material i the followig referece: Moder Cotrol Sytem by Richard C Dorf ad Robert H Bihop, Pretice Hall

More information

5.80 Small-Molecule Spectroscopy and Dynamics

5.80 Small-Molecule Spectroscopy and Dynamics MIT OpeCurseWare http://cw.mit.edu 5.8 Small-Mlecule Spectrscpy ad Dyamics Fall 8 Fr ifrmati abut citig these materials r ur Terms f Use, visit: http://cw.mit.edu/terms. 5.8 Lecture #33 Fall, 8 Page f

More information

A Tail Bound For Sums Of Independent Random Variables And Application To The Pareto Distribution

A Tail Bound For Sums Of Independent Random Variables And Application To The Pareto Distribution Applied Mathematic E-Note, 9009, 300-306 c ISSN 1607-510 Available free at mirror ite of http://wwwmaththuedutw/ ame/ A Tail Boud For Sum Of Idepedet Radom Variable Ad Applicatio To The Pareto Ditributio

More information

REVIEW OF SIMPLE LINEAR REGRESSION SIMPLE LINEAR REGRESSION

REVIEW OF SIMPLE LINEAR REGRESSION SIMPLE LINEAR REGRESSION REVIEW OF SIMPLE LINEAR REGRESSION SIMPLE LINEAR REGRESSION I liear regreio, we coider the frequecy ditributio of oe variable (Y) at each of everal level of a ecod variable (X). Y i kow a the depedet variable.

More information

D.S.G. POLLOCK: TOPICS IN TIME-SERIES ANALYSIS STATISTICAL FOURIER ANALYSIS

D.S.G. POLLOCK: TOPICS IN TIME-SERIES ANALYSIS STATISTICAL FOURIER ANALYSIS STATISTICAL FOURIER ANALYSIS The Furier Represetati f a Sequece Accrdig t the basic result f Furier aalysis, it is always pssible t apprximate a arbitrary aalytic fucti defied ver a fiite iterval f the

More information

5.1 Two-Step Conditional Density Estimator

5.1 Two-Step Conditional Density Estimator 5.1 Tw-Step Cditial Desity Estimatr We ca write y = g(x) + e where g(x) is the cditial mea fucti ad e is the regressi errr. Let f e (e j x) be the cditial desity f e give X = x: The the cditial desity

More information

M6c: Design of Stable Open. Channels. future conditions, reasonable cost, minimal. These objectives must be met considering

M6c: Design of Stable Open. Channels. future conditions, reasonable cost, minimal. These objectives must be met considering M6c: Deig f Stale Ope Chael Adequate cveyace capacity Stale chael Prvide aquatic life haitat Thee jective mut e met ciderig future cditi, reaale ct, miimal lad cumpti, ad afety. Trapezidal Secti (Figure

More information

State space systems analysis

State space systems analysis State pace ytem aalyi Repreetatio of a ytem i tate-pace (tate-pace model of a ytem To itroduce the tate pace formalim let u tart with a eample i which the ytem i dicuio i a imple electrical circuit with

More information

Solutions to Midterm II. of the following equation consistent with the boundary condition stated u. y u x y

Solutions to Midterm II. of the following equation consistent with the boundary condition stated u. y u x y Sltis t Midterm II Prblem : (pts) Fid the mst geeral slti ( f the fllwig eqati csistet with the bdary cditi stated y 3 y the lie y () Slti : Sice the system () is liear the slti is give as a sperpsiti

More information

x 2 x 3 x b 0, then a, b, c log x 1 log z log x log y 1 logb log a dy 4. dx As tangent is perpendicular to the x axis, slope

x 2 x 3 x b 0, then a, b, c log x 1 log z log x log y 1 logb log a dy 4. dx As tangent is perpendicular to the x axis, slope The agle betwee the tagets draw t the parabla y = frm the pit (-,) 5 9 6 Here give pit lies the directri, hece the agle betwee the tagets frm that pit right agle Ratig :EASY The umber f values f c such

More information

Statistics and Chemical Measurements: Quantifying Uncertainty. Normal or Gaussian Distribution The Bell Curve

Statistics and Chemical Measurements: Quantifying Uncertainty. Normal or Gaussian Distribution The Bell Curve Statitic ad Chemical Meauremet: Quatifyig Ucertaity The bottom lie: Do we trut our reult? Should we (or ayoe ele)? Why? What i Quality Aurace? What i Quality Cotrol? Normal or Gauia Ditributio The Bell

More information

On the 2-Domination Number of Complete Grid Graphs

On the 2-Domination Number of Complete Grid Graphs Ope Joural of Dicrete Mathematic, 0,, -0 http://wwwcirporg/oural/odm ISSN Olie: - ISSN Prit: - O the -Domiatio Number of Complete Grid Graph Ramy Shahee, Suhail Mahfud, Khame Almaea Departmet of Mathematic,

More information

On the Signed Domination Number of the Cartesian Product of Two Directed Cycles

On the Signed Domination Number of the Cartesian Product of Two Directed Cycles Ope Joural of Dicrete Mathematic, 205, 5, 54-64 Publihed Olie July 205 i SciRe http://wwwcirporg/oural/odm http://dxdoiorg/0426/odm2055005 O the Siged Domiatio Number of the Carteia Product of Two Directed

More information

Tools Hypothesis Tests

Tools Hypothesis Tests Tool Hypothei Tet The Tool meu provide acce to a Hypothei Tet procedure that calculate cofidece iterval ad perform hypothei tet for mea, variace, rate ad proportio. It i cotrolled by the dialog box how

More information

Heat Equation: Maximum Principles

Heat Equation: Maximum Principles Heat Equatio: Maximum Priciple Nov. 9, 0 I thi lecture we will dicu the maximum priciple ad uiquee of olutio for the heat equatio.. Maximum priciple. The heat equatio alo ejoy maximum priciple a the Laplace

More information

Sound Absorption Characteristics of Membrane- Based Sound Absorbers

Sound Absorption Characteristics of Membrane- Based Sound Absorbers Purdue e-pubs Publicatis f the Ray W. Schl f Mechaical Egieerig 8-28-2003 Sud Absrpti Characteristics f Membrae- Based Sud Absrbers J Stuart Blt, blt@purdue.edu Jih Sg Fllw this ad additial wrks at: http://dcs.lib.purdue.edu/herrick

More information

, the random variable. and a sample size over the y-values 0:1:10.

, the random variable. and a sample size over the y-values 0:1:10. Lecture 3 (4//9) 000 HW PROBLEM 3(5pts) The estimatr i (c) f PROBLEM, p 000, where { } ~ iid bimial(,, is 000 e f the mst ppular statistics It is the estimatr f the ppulati prprti I PROBLEM we used simulatis

More information

Erick L. Oberstar Fall 2001 Project: Sidelobe Canceller & GSC 1. Advanced Digital Signal Processing Sidelobe Canceller (Beam Former)

Erick L. Oberstar Fall 2001 Project: Sidelobe Canceller & GSC 1. Advanced Digital Signal Processing Sidelobe Canceller (Beam Former) Erick L. Obertar Fall 001 Project: Sidelobe Caceller & GSC 1 Advaced Digital Sigal Proceig Sidelobe Caceller (Beam Former) Erick L. Obertar 001 Erick L. Obertar Fall 001 Project: Sidelobe Caceller & GSC

More information

Mean residual life of coherent systems consisting of multiple types of dependent components

Mean residual life of coherent systems consisting of multiple types of dependent components Mea residual life f cheret systems csistig f multiple types f depedet cmpets Serka Eryilmaz, Frak P.A. Cle y ad Tahai Cle-Maturi z February 20, 208 Abstract Mea residual life is a useful dyamic characteristic

More information

MATHEMATICS 9740/01 Paper 1 14 Sep hours

MATHEMATICS 9740/01 Paper 1 14 Sep hours Cadidate Name: Class: JC PRELIMINARY EXAM Higher MATHEMATICS 9740/0 Paper 4 Sep 06 3 hurs Additial Materials: Cver page Aswer papers List f Frmulae (MF5) READ THESE INSTRUCTIONS FIRST Write yur full ame

More information

Existence theory for sequential fractional differential equations with anti-periodic type boundary conditions

Existence theory for sequential fractional differential equations with anti-periodic type boundary conditions Ope Math. 26; 4: 723 735 Ope Mathematic Ope Acce Reearch Article Mhammed H. Aqla, Ahmed Alaedi, Bahir Ahmad*, ad Jua J. Niet Exitece thery fr equetial fractial differetial equati with ati-peridic type

More information

A tail bound for sums of independent random variables : application to the symmetric Pareto distribution

A tail bound for sums of independent random variables : application to the symmetric Pareto distribution A tail boud for um of idepedet radom variable : applicatio to the ymmetric Pareto ditributio Chritophe Cheeau To cite thi verio: Chritophe Cheeau. A tail boud for um of idepedet radom variable : applicatio

More information

UNIVERSITY OF TECHNOLOGY. Department of Mathematics PROBABILITY THEORY, STATISTICS AND OPERATIONS RESEARCH GROUP. Memorandum COSOR 76-10

UNIVERSITY OF TECHNOLOGY. Department of Mathematics PROBABILITY THEORY, STATISTICS AND OPERATIONS RESEARCH GROUP. Memorandum COSOR 76-10 EI~~HOVEN UNIVERSITY OF TECHNOLOGY Departmet f Mathematics PROBABILITY THEORY, STATISTICS AND OPERATIONS RESEARCH GROUP Memradum COSOR 76-10 O a class f embedded Markv prcesses ad recurrece by F.H. Sims

More information

x z Increasing the size of the sample increases the power (reduces the probability of a Type II error) when the significance level remains fixed.

x z Increasing the size of the sample increases the power (reduces the probability of a Type II error) when the significance level remains fixed. ] z-tet for the mea, μ If the P-value i a mall or maller tha a pecified value, the data are tatitically igificat at igificace level. Sigificace tet for the hypothei H 0: = 0 cocerig the ukow mea of a populatio

More information

Explicit scheme. Fully implicit scheme Notes. Fully implicit scheme Notes. Fully implicit scheme Notes. Notes

Explicit scheme. Fully implicit scheme Notes. Fully implicit scheme Notes. Fully implicit scheme Notes. Notes Explicit cheme So far coidered a fully explicit cheme to umerically olve the diffuio equatio: T + = ( )T + (T+ + T ) () with = κ ( x) Oly table for < / Thi cheme i ometime referred to a FTCS (forward time

More information

[1 & α(t & T 1. ' ρ 1

[1 & α(t & T 1. ' ρ 1 NAME 89.304 - IGNEOUS & METAMORPHIC PETROLOGY DENSITY & VISCOSITY OF MAGMAS I. Desity The desity (mass/vlume) f a magma is a imprtat parameter which plays a rle i a umber f aspects f magma behavir ad evluti.

More information

Performance-Based Plastic Design (PBPD) Procedure

Performance-Based Plastic Design (PBPD) Procedure Performace-Baed Platic Deig (PBPD) Procedure 3. Geeral A outlie of the tep-by-tep, Performace-Baed Platic Deig (PBPD) procedure follow, with detail to be dicued i ubequet ectio i thi chapter ad theoretical

More information

Increasing Voltage Gain by New Structure of Inductive Switching DC-DC Converter

Increasing Voltage Gain by New Structure of Inductive Switching DC-DC Converter AUT Jural f Electrical Egeerg AUT J. Elec. Eg., 49((0773-78 DO: 0.060/eej.07.555.4978 creag Vltage Ga by New Structure f ductie Switchg D-D erter S. Nabati, A. Siadata *, S. B. Mzafari Departmet f Electrical

More information

Brief Review of Linear System Theory

Brief Review of Linear System Theory Brief Review of Liear Sytem heory he followig iformatio i typically covered i a coure o liear ytem theory. At ISU, EE 577 i oe uch coure ad i highly recommeded for power ytem egieerig tudet. We have developed

More information

Cryptographic Hash Functions and Message Authentication Codes. Reading: Chapter 4 of Katz & Lindell

Cryptographic Hash Functions and Message Authentication Codes. Reading: Chapter 4 of Katz & Lindell Cryptographic Hah Fuctio ad Meage Autheticatio Code Readig: Chapter 4 of Katz & Lidell 1 Hah fuctio A fuctio mappig from a domai to a maller rage (thu ot ijective). Applicatio: Fat looup (hah table) Error

More information

K [f(t)] 2 [ (st) /2 K A GENERALIZED MEIJER TRANSFORMATION. Ku(z) ()x) t -)-I e. K(z) r( + ) () (t 2 I) -1/2 e -zt dt, G. L. N. RAO L.

K [f(t)] 2 [ (st) /2 K A GENERALIZED MEIJER TRANSFORMATION. Ku(z) ()x) t -)-I e. K(z) r( + ) () (t 2 I) -1/2 e -zt dt, G. L. N. RAO L. Iterat. J. Math. & Math. Scl. Vl. 8 N. 2 (1985) 359-365 359 A GENERALIZED MEIJER TRANSFORMATION G. L. N. RAO Departmet f Mathematics Jamshedpur C-perative Cllege f the Rachi Uiversity Jamshedpur, Idia

More information

Chapter 9. Key Ideas Hypothesis Test (Two Populations)

Chapter 9. Key Ideas Hypothesis Test (Two Populations) Chapter 9 Key Idea Hypothei Tet (Two Populatio) Sectio 9-: Overview I Chapter 8, dicuio cetered aroud hypothei tet for the proportio, mea, ad tadard deviatio/variace of a igle populatio. However, ofte

More information

Note 8 Root-Locus Techniques

Note 8 Root-Locus Techniques Lecture Nte f Ctrl Syte I - ME 43/Alyi d Sythei f Lier Ctrl Syte - ME862 Nte 8 Rt-Lcu Techique Deprtet f Mechicl Egieerig, Uiverity Of Sktchew, 57 Cpu Drive, Skt, S S7N 5A9, Cd Lecture Nte f Ctrl Syte

More information

COMPUTING CONFIDENCE INTERVALS FOR OUTPUT ORIENTED DEA MODELS: AN APPLICATION TO AGRICULTURAL RESEARCH IN BRAZIL

COMPUTING CONFIDENCE INTERVALS FOR OUTPUT ORIENTED DEA MODELS: AN APPLICATION TO AGRICULTURAL RESEARCH IN BRAZIL COMPUTING CONFIDENCE INTERVALS FOR OUTPUT ORIENTED DEA MODELS: AN APPLICATION TO AGRICULTURAL RESEARCH IN BRAZIL Gerald da Silva e Suza Miria Oliveira de Suza Eliae Gçalves Gmes Brazilia Agricultural Research

More information

While established performance

While established performance KATHRYN WILKENS i aitat prfer f fiace i the Departet f Maageet at Wrceter Plytechic Ititute i Wrceter, MA. JOE ZHU i aitat prfer f perati i the Departet f Maageet at Wrceter Plytechic Ititute i Wrceter,

More information

Estimation of Monthly Average Hourly Global Solar Radiation from the Daily Value in Çanakkale, Turkey

Estimation of Monthly Average Hourly Global Solar Radiation from the Daily Value in Çanakkale, Turkey Jural f Clea Eergy Techlgie, Vl. 5, N. 5, September 017 Etimati f Mthly Average Hurly Glbal Slar Radiati frm the Daily Value i Çaakkale, Turkey Özge Ayvazğluyükel ad Ümmüha Başara Filik lt f mdel perfrmed

More information

w (1) ˆx w (1) x (1) /ρ and w (2) ˆx w (2) x (2) /ρ.

w (1) ˆx w (1) x (1) /ρ and w (2) ˆx w (2) x (2) /ρ. 2 5. Weighted umber of late jobs 5.1. Release dates ad due dates: maximimizig the weight of o-time jobs Oce we add release dates, miimizig the umber of late jobs becomes a sigificatly harder problem. For

More information

STUDENT S t-distribution AND CONFIDENCE INTERVALS OF THE MEAN ( )

STUDENT S t-distribution AND CONFIDENCE INTERVALS OF THE MEAN ( ) STUDENT S t-distribution AND CONFIDENCE INTERVALS OF THE MEAN Suppoe that we have a ample of meaured value x1, x, x3,, x of a igle uow quatity. Aumig that the meauremet are draw from a ormal ditributio

More information

Name Student ID. A student uses a voltmeter to measure the electric potential difference across the three boxes.

Name Student ID. A student uses a voltmeter to measure the electric potential difference across the three boxes. Name Student ID II. [25 pt] Thi quetin cnit f tw unrelated part. Part 1. In the circuit belw, bulb 1-5 are identical, and the batterie are identical and ideal. Bxe,, and cntain unknwn arrangement f linear

More information

Economics of a reservation system for morning commute

Economics of a reservation system for morning commute Paper ubmitted t the ITEA Aual Cferece Traprtati Ecmic (Kuhm ectar), 04 Ecmic f a reervati ytem fr mrig cmmute Wei Liu, Hai Yag ad Fagi Zhag Departmet f Civil ad Evirmetal Egieerig, The Hg Kg Uiverity

More information

1. Introduction: A Mixing Problem

1. Introduction: A Mixing Problem CHAPTER 7 Laplace Tranfrm. Intrductin: A Mixing Prblem Example. Initially, kg f alt are dilved in L f water in a tank. The tank ha tw input valve, A and B, and ne exit valve C. At time t =, valve A i pened,

More information

TESTS OF SIGNIFICANCE

TESTS OF SIGNIFICANCE TESTS OF SIGNIFICANCE Seema Jaggi I.A.S.R.I., Library Aveue, New Delhi eema@iari.re.i I applied ivetigatio, oe i ofte itereted i comparig ome characteritic (uch a the mea, the variace or a meaure of aociatio

More information

The generation of successive approximation methods for Markov decision processes by using stopping times

The generation of successive approximation methods for Markov decision processes by using stopping times The geerati f successive apprximati methds fr Markv decisi prcesses by usig stppig times Citati fr published versi (APA): va Nue, J. A. E. E., & Wessels, J. (1976). The geerati f successive apprximati

More information

A criterion for easiness of certain SAT-problems

A criterion for easiness of certain SAT-problems A criterio for eaie of certai SAT-problem Berd R. Schuh Dr. Berd Schuh, D-50968 Köl, Germay; berd.chuh@etcologe.de keyword: compleity, atifiability, propoitioal logic, P, NP, -i-3sat, eay/hard itace Abtract.

More information

Phys 2310 Wed. Oct. 4, 2017 Today s Topics

Phys 2310 Wed. Oct. 4, 2017 Today s Topics Phy 30 Wed. Oct. 4, 07 Tday Tpic Ctiue Chapter 33: Gemetric Optic Readig r Next Time By Mday: Readig thi Week Fiih Ch. 33 Lee, Mirrr ad Prim Hmewrk Due Oct., 07 Y&F Ch. 3: #3., 3.5 Ch. 33: #33.3, 33.7,

More information

Active redundancy allocation in systems. R. Romera; J. Valdés; R. Zequeira*

Active redundancy allocation in systems. R. Romera; J. Valdés; R. Zequeira* Wrkig Paper -6 (3) Statistics ad Ecmetrics Series March Departamet de Estadística y Ecmetría Uiversidad Carls III de Madrid Calle Madrid, 6 893 Getafe (Spai) Fax (34) 9 64-98-49 Active redudacy allcati

More information

Study of Energy Eigenvalues of Three Dimensional. Quantum Wires with Variable Cross Section

Study of Energy Eigenvalues of Three Dimensional. Quantum Wires with Variable Cross Section Adv. Studies Ther. Phys. Vl. 3 009. 5 3-0 Study f Eergy Eigevalues f Three Dimesial Quatum Wires with Variale Crss Secti M.. Sltai Erde Msa Departmet f physics Islamic Aad Uiversity Share-ey rach Ira alrevahidi@yah.cm

More information

Chapter 9 Compressible Flow 667

Chapter 9 Compressible Flow 667 Chapter 9 Cmpreible Flw 667 9.57 Air flw frm a tank thrugh a nzzle int the tandard atmphere, a in Fig. P9.57. A nrmal hck tand in the exit f the nzzle, a hwn. Etimate (a) the tank preure; and (b) the ma

More information

Pipe Networks - Hardy Cross Method Page 1. Pipe Networks

Pipe Networks - Hardy Cross Method Page 1. Pipe Networks Pie Netwrks - Hardy Crss etd Page Pie Netwrks Itrducti A ie etwrk is a itercected set f ies likig e r mre surces t e r mre demad (delivery) its, ad ca ivlve ay umber f ies i series, bracig ies, ad arallel

More information

Physical Chemistry Laboratory I CHEM 445 Experiment 2 Partial Molar Volume (Revised, 01/13/03)

Physical Chemistry Laboratory I CHEM 445 Experiment 2 Partial Molar Volume (Revised, 01/13/03) Physical Chemistry Labratry I CHEM 445 Experimet Partial Mlar lume (Revised, 0/3/03) lume is, t a gd apprximati, a additive prperty. Certaily this apprximati is used i preparig slutis whse ccetratis are

More information

Chapter 8. Root Locus Techniques

Chapter 8. Root Locus Techniques Chapter 8 Rt Lcu Technique Intrductin Sytem perfrmance and tability dt determined dby cled-lp l ple Typical cled-lp feedback cntrl ytem G Open-lp TF KG H Zer -, - Ple 0, -, -4 K 4 Lcatin f ple eaily fund

More information

MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.265/15.070J Fall 2013 Lecture 16 11/04/2013. Ito integral. Properties

MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.265/15.070J Fall 2013 Lecture 16 11/04/2013. Ito integral. Properties MASSACHUSES INSIUE OF ECHNOLOGY 6.65/15.7J Fall 13 Lecture 16 11/4/13 Ito itegral. Propertie Cotet. 1. Defiitio of Ito itegral. Propertie of Ito itegral 1 Ito itegral. Exitece We cotiue with the cotructio

More information

6.3 Testing Series With Positive Terms

6.3 Testing Series With Positive Terms 6.3. TESTING SERIES WITH POSITIVE TERMS 307 6.3 Testig Series With Positive Terms 6.3. Review of what is kow up to ow I theory, testig a series a i for covergece amouts to fidig the i= sequece of partial

More information

Roger Apéry's proof that zeta(3) is irrational

Roger Apéry's proof that zeta(3) is irrational Cliff Bott cliffbott@hotmail.com 11 October 2011 Roger Apéry's proof that zeta(3) is irratioal Roger Apéry developed a method for searchig for cotiued fractio represetatios of umbers that have a form such

More information

Comparative analysis of bayesian control chart estimation and conventional multivariate control chart

Comparative analysis of bayesian control chart estimation and conventional multivariate control chart America Jural f Theretical ad Applied Statistics 3; ( : 7- ublished lie Jauary, 3 (http://www.sciecepublishiggrup.cm//atas di:.648/.atas.3. Cmparative aalysis f bayesia ctrl chart estimati ad cvetial multivariate

More information

Last time: Completed solution to the optimum linear filter in real-time operation

Last time: Completed solution to the optimum linear filter in real-time operation 6.3 tochatic Etimatio ad Cotrol, Fall 4 ecture at time: Completed olutio to the oimum liear filter i real-time operatio emi-free cofiguratio: t D( p) F( p) i( p) dte dp e π F( ) F( ) ( ) F( p) ( p) 4444443

More information

Statistical Inference Procedures

Statistical Inference Procedures Statitical Iferece Procedure Cofidece Iterval Hypothei Tet Statitical iferece produce awer to pecific quetio about the populatio of iteret baed o the iformatio i a ample. Iferece procedure mut iclude a

More information

8.6 Order-Recursive LS s[n]

8.6 Order-Recursive LS s[n] 8.6 Order-Recurive LS [] Motivate ti idea wit Curve Fittig Give data: 0,,,..., - [0], [],..., [-] Wat to fit a polyomial to data.., but wic oe i te rigt model?! Cotat! Quadratic! Liear! Cubic, Etc. ry

More information