Best response equivalence

Size: px
Start display at page:

Download "Best response equivalence"

Transcription

1 Gmes nd Economc Behvor ) Best response equvlence Stephen Morrs,, Tksh U b Deprtment of Economcs, Yle Unversty, USA b Fculty of Economcs, Yokohm Ntonl Unversty, Jpn Receved 2 August 2002 Avlble onlne 15 Aprl 2004 Abstrct Two gmes re best-response equvlent f they hve the sme best-response correspondence. We provde chrcterzton of when two gmes re best-response equvlent. The chrcterztons explot dul reltonshp between pyoff dfferences nd belefs. Some potentl gme rguments [Gmes Econ. Behv ) 124] rely only on the property tht potentl gmes re best-response equvlent to dentcl nterest gmes. Our results show tht lrge clss of gmes re best-response equvlent to dentcl nterest gmes, but re not potentl gmes. Thus we show how some exstng potentl gme rguments cn be extended Elsever Inc. All rghts reserved. JEL clssfcton: C72 Keywords: Best response equvlence; Dulty; Frks Lemm; Potentl gmes 1. Introducton We consder three progressvely stronger equvlence reltons on gmes nd chrcterze ech of them. Two gmes re best-response equvlent f they hve the sme best-response correspondence. Two gmes re better-response equvlent f, for every pr of strteges, they gree when one strtegy s better thn the other. * Correspondng uthor s ddress: Cowles Foundton, P.O. Box , New Hven, CT , USA. E-ml ddresses: stephen.morrs@yle.edu S. Morrs), ou@ynu.c.jp T. U) /$ see front mtter 2004 Elsever Inc. All rghts reserved. do: /j.geb

2 S. Morrs, T. U / Gmes nd Economc Behvor ) Two gmes re von Neumnn Morgenstern equvlent VNM-equvlent) f, for ech plyer, the pyoff functon n one gme s equl to constnt tmes the pyoff functon n the other gme, plus functon tht depends only on the opponents strteges. Two gmes re VNM-equvlent f nd only f, for ech plyer, there s constnt w > 0 such tht the rto of pyoff dfferences from swtchng between one strtegy to nother strtegy s lwys w. The constnt w s thus ndependent of the strteges beng compred. Two gmes re better-response equvlent f nd only f they hve the sme domnnce reltons nd, for ech plyer nd ech pr of strteges nd such tht nether strtegy strctly domntes the other, there exsts constnt w > 0 such tht the rto of pyoff dfferences from swtchng between nd s lwys w. In generl, ths s weker requrement thn VNM-equvlence. It s weker both becuse the proportonl pyoff dfferences property s no longer requred to hold between some strtegy prs, nd becuse the weght w s not necessrly ndependent of the strtegy pr. But f the gme does not hve domnted strteges, the weghts cn no longer depend on the strteges beng compred, nd better-response equvlence collpses to VNM-equvlence. Two gmes re best-response equvlent f nd only f, for ech plyer nd ech pr of strteges nd such tht both strteges re best response to some belef, there exsts constnt w > 0 such tht the rto of pyoff dfferences from swtchng between nd s lwys w. Even f gme hs no domnted strteges, ths s weker requrement thn VNM-equvlence. In gmes wth dmnshng mrgnl returns, best-response equvlence s lwys strctly weker requrement thn VNM-equvlence. Exmples re gven n the pper. The most extensve dscusson nd pplctons of these reltons hve come n the lterture on potentl gmes. Monderer nd Shpley 1996b) sd tht gme ws potentl gme f there exsts potentl functon, defned on the strtegy spce, wth the property tht the chnge n ny plyer s pyoff functon from swtchng between ny two of hs strteges holdng other plyers strteges fxed) ws equl to the chnge n the potentl functon. 1 A gme s weghted potentl gme, f the pyoff chnges re proportonl for ech plyer. Thus gme s weghted potentl gme f nd only f t s VNM-equvlent to gme wth dentcl pyoff functons. Whle some results usng potentl or weghted potentl gme rguments re usng the VNM-equvlence to dentcl nterest gmes, other rguments re just usng the better-response equvlence nd even only best-response equvlence mplctons of VNM-equvlence. 2 Any pper tht dels only wth equlbr s usng only best-response equvlence e.g., Neymn, 1997; U, 2001; Morrs nd U, 2002). Smlrly, fcttous ply only uses the best-response 1 See lso U 2000) for chrcterzton nd exmples of potentl gmes. 2 Arguments tht explot potentl rguments to prove the exstence of pure strtegy equlbrum e.g., Rosenthl, 1973) only use ordnl propertes of pyoffs. Monderer nd Shpley 1996b) ntroduced ordnl potentl gmes nd Voorneveld 2000) nd Dubey et l. 2002) showed how ordnl potentl gmes cn be wekened to only requre pure strtegy best-response equvlence.

3 262 S. Morrs, T. U / Gmes nd Economc Behvor ) propertes of the gme Monderer nd Shpley, 1996). 3 An pplcton usng only better-response equvlence but not the VNM-equvlence ppers n Morrs 1999). Some ppers studyng quntl responses or stochstc best responses n potentl gmes use the full power of VNM-equvlence e.g., Blume, 1993; Brock nd Durluf, 2001; Anderson et l., 2001; U, 2002). 4 The fct tht VNM-equvlence s the sme s better-response equvlence n the bsence of domnted strteges nd my be dfferent n the presence of domnted strteges hs been noted n number of contexts see Sel, 1992; Blume, 1993, p. 409; Monderer nd Shpley, 1996b, footnote 9; Mskn nd Trole, 2001, p. 209). However, our chrcterztons of better-response equvlence n the presence of domnted strteges nd of the sgnfcnt gp between better-response equvlence nd best-response equvlence fll gp n the lterture. 5 The pper s orgnzed s follows. In Secton 2, we descrbe our notons of equvlence nd gve n exmple llustrtng the dfferences. In Secton 3, we report our chrcterztons. In Secton 4, we restrct ttenton to clss of gmes where best-response equvlence s strctly weker requrement thn VNM-equvlence nd chrcterze tht clss. We lso dscuss n extenson to gmes wth nfnte strtegy spces nd ts pplcton. Secton 5 brefly dscusses better-response nd best-response equvlence n the mxed strtegy extenson of gme. 2. Equvlence propertes of gmes A gme conssts of fnte set of plyers N nd fnte strtegy set A for N, nd pyoff functon g : A R for N where A = N A. We wrte A = j A j nd = j ) j A. We smply denote gme by g = g ) N. Throughout the pper, we regrd g, ) : A R s vector n R A. We wrte g, ) g, ) f g, )>g, ) for ll A,ndg, ) g, ) f g, ) g, ) for ll A. For N, let A ) denote the set of ll probblty dstrbutons over A. We cll ech element of A ) plyer s belef. For X A,letΛ,X g ) A ) be set of plyer s belefs such tht plyer wth pyoff functon g nd belef λ Λ,X g ) wekly prefers to ny strtegy n X : Λ,X g ) { } )) = λ A ) λ ) g, ) g, 0forll X. When X s sngleton,.e., X ={ }, we wrte Λ, g ) nsted of Λ, { } g ). 3 Sel 1999) estblshes convergence of fcttous ply n clss of one-gnst-ll gmes. These gmes re best-response equvlent to dentcl nterest gmes, but not potentl gmes. 4 More precsely, they use the full power of VNM-equvlence such tht the constnt w s the sme for ll the plyers. 5 Mertens 1987) studed vrous notons of best-response equvlence, but wth hs more bstrct strtegy spces nd focus on dmssble best responses, there s lttle overlp wth the mterl n ths pper.

4 S. Morrs, T. U / Gmes nd Economc Behvor ) We re nterested n chrcterzng two equvlence reltons on gmes cptured by these sets of belefs by whch plyers prefer one prtculr strtegy. Defnton 1. Agmeg s better-response equvlent to g = g ) N f, for ech N, Λ, g ) = Λ, ) g for ll, A. Defnton 2. Agmeg s best-response equvlent to g = g ) N f, for ech N, Λ,A g ) = Λ,A g ) for ll A. If g s better-response equvlent to g,theng s best-response equvlent to g,snce Λ,A g ) = Λ, g ). A An esy suffcent condton for better-response equvlence s the followng. 6 Defnton 3. Agmeg s VNM-equvlent to g = g ) N f, for ech N, there exsts postve constnt w > 0 nd functon Q : A R such tht g, ) = w g, ) + Q ) for ll A. It s strghtforwrd to see tht f g s VNM-equvlent to g,then g, ) g, ) = w g, ) g, )) for ll, A. Conversely, f ths s true, then functon Q : A R such tht Q ) = g, ) w g, ) s well defned, nd thus g s VNM-equvlent to g. Thus, we hve the followng lemm. Lemm 1. A gme g s VNM-equvlent to g f nd only f, for ech N, there exsts w such tht g, ) g, ) = w g, ) g, )) 1) for ll, A. It s strghtforwrd to see tht VNM-equvlence s suffcent for better-response equvlence. In fct, 1) mples tht 6 Blume 1993) clled ths property strongly best-response equvlent.

5 264 S. Morrs, T. U / Gmes nd Economc Behvor ) )) λ ) g, ) g, = w λ ) g, ) g )), for ll λ A ) nd thus Λ, g ) = Λ, g ) for ll, A. Best-response, better-response, nd VNM-equvlence re equvlence reltons. Thus, they defne n equvlence clss of gmes. For exmple, weghted potentl gmes Monderer nd Shpley, 1996b) wth weghted potentl functon f : A R re regrded s VNM-equvlence clss of n dentcl nterest gme f = f ) N wth f = f for ll N. Ths s cler by Lemm 1 nd the followng orgnl defnton of weghted potentl gmes. Defnton 4. Agmeg = g ) N s weghted potentl gme f there exsts weghted potentl functon f : A R nd w > 0 for ech N such tht g, ) g, ) = w f, ) f, )) for ll, A.Ifw = 1forll N, g s clled potentl gme nd f s clled potentl functon. As the concept of VNM-equvlence leds us to the defnton of weghted potentl gmes, the concept of better-response equvlence nd tht of best-response equvlence led us to the followng defntons of new clsses of gmes. Defnton 5. Agmeg = g ) N s better-response potentl gme f t s betterresponse equvlent to n dentcl nterest gme f = f ) N wth f = f for ll N. A functon f s clled better-response potentl functon. Defnton 6. Agmeg = g ) N s best-response potentl gme f t s best-response equvlent to n dentcl nterest gme f = f ) N wth f = f for ll N. A functon f s clled best-response potentl functon. Voorneveld 2000) clled gme best-response potentl gme f ts best-response correspondence concdes wth tht of n dentcl nterest gme over the clss of belefs such tht λ ) = 0 or 1. Thus, best-response potentl gmes n ths pper form specl clss of those n Voorneveld 2000). Exstng potentl gme results tht rely only on better-response equvlence or bestresponse equvlence, such s those mentoned n the ntroducton, utomtclly hold for the lrger clss of better-response potentl gmes or tht of best-response potentl gmes. Thus, we re nterested n exctly when nd to wht extent better-response nd best-response equvlence re weker requrements thn VNM-equvlence. Notce tht best-response nd better-response equvlence re clerly weker requrements thn VNM-equvlence, becuse the ltter mposes too mny constrnts on pyoffs from domnted strteges. Moreover, best-response equvlence s sgnfcntly weker thn better-response equvlence, s shown by the followng exmple.

6 S. Morrs, T. U / Gmes nd Economc Behvor ) Consder two plyer, three strtegy, symmetrc pyoff gme gx, y) prmeterzed by x, y) R 2 ++, where ech plyer s pyoffs re gven by the followng pyoff mtrx where the plyer s own strteges re represented by rows nd hs opponent s strteges re represented by columns): x x 2x y y y. In the specl cse where x = y = 1, we hve gme g1, 1) wth the followng pyoff mtrx: If row plyer hs belef λ k) = π k for k {1, 2, 3}, he prefers strtegy 1 to strtegy 2 f nd only f π 1 π 2 + 2π 3 ; he prefers strtegy 1 to strtegy 3 f nd only f x + 2y)π 1 x y)π 2 + 2x + y)π 3 ; he prefers strtegy 3 to strtegy 2 f nd only f π 3 π 2 + 2π 1. Thus the regon of ndfference between strteges 1 nd 2, nd between strtegy 2 nd 3, does not depend on x nd y. Moreover, whenever strtegy 1 or 3) s preferred to strtegy 2, t s lso preferred to strtegy 3 or 1). Thus the best response regons for ths gme re s n Fg. 1, for ny x, y) R Thus gx, y) s best-response equvlent to g1, 1) for ny Fg. 1. The best response regons.

7 266 S. Morrs, T. U / Gmes nd Economc Behvor ) x, y) R On the other hnd, the regon of ndfference between strteges 1 nd 3 does depend on x nd y: n prtculr, gx, y) s better-response equvlent to g1, 1) f nd only f x = y. We wll dscuss ths exmple gn n Secton Results 3.1. Generc propertes of gmes We wll ppel to some generc propertes of gmes,.e., propertes tht wll hold for ll but Lebesgue mesure zero set of pyoffs s long s ech plyer hs t lest two ctons). G1. For ll N, fg, ) g, ),theng, ) g, ) for dstnct, A. G2. For ll N, vectors g, ) g, ) nd g, ) g, ) re lnerly ndependent for dstnct,, A. G3. For ll N, fλ,a g ) Λ,A g ),then Λ,A \ { } g ) \Λ, g ) for dstnct, A Better-response equvlence Strtegy strctly domntes n gme g we wrte g )fg, ) g, ), or, equvlently, Λ, g ) =. Strteges nd re better-response comprble we wrte g ) f nether g nor g. Proposton 1. If gmes g nd g stsfy generc property G1, then g s better-response equvlent to g f nd only f, for ech N, ) they hve the sme domnnce reltons g = g ) nd b) whenever s better-response comprble to g ), there exsts w, )>0 such tht g, ) g, ) = w, ) g, ) g, )). 2) Frks Lemm 7 plys centrl role n the proofs. Lemm 2 Frks Lemm). For vectors 0, 1,..., m R n, the followng two condtons re equvlent. If 1.y),..., m.y) 0 for y R n,then 0.y) 0. 8 There exsts x 1,...,x m 0 such tht x x m m = 0. 7 See textbook of convex nlyss such s the recent one by Hrrt-Urruty nd Lemréchl 2001), or the clssc one by Rockfellr 1970). 8 I.e., f n j=1 j y j 0 for ech = 1,...,m,then n j=1 0j y j 0.

8 S. Morrs, T. U / Gmes nd Economc Behvor ) Proof of Proposton 1. We frst show tht ) nd b) re suffcent for the better-response equvlence of g nd g.if g, then b) mples tht )) λ ) g, ) g, = w, ) g, ) g )), λ ) nd thus Λ, g ) = Λ, ) g. If g,then Λ, g ) = Λ, ) g = A ). If g,then Λ, g ) = Λ, ) g =. To prove necessty, suppose tht g s better-response equvlent to g.snce Λ, g ) = Λ, ) g, we hve g ) Λ, g = Λ, g ) = g nd thus ) holds. To prove b), suppose tht g. We know tht g.letλ A ) be such tht )) λ ) g, ) g, 0. Snce λ Λ, g ) = Λ, g ), λ ) g, ) g )), 0. Ths mples tht f y ) A R A s such tht )) g, ) g, 0, then y y 0 forll A, y g, ) g )), 0. By Frks Lemm, there exst x 0ndz 0for A such tht x g, ) g, )) z δ ) = g, ) g, ))

9 268 S. Morrs, T. U / Gmes nd Economc Behvor ) where δ : A R s such tht δ ) = 1f = nd δ ) = 0otherwse. Thus, x g, ) g, )) g, ) g, ). If x = 0, then g, ) g, ) 0. However, ths s mpossble snce g mples tht does not strctly domnte n g nd G1 requres tht f does not strctly domnte, then t s not the cse tht g, ) g, ) 0. Thus, x > 0. Symmetrclly, we hve x g, ) g, ) ) g, ) g, ) where x > 0. Thus, ) x x g, ) g, )) 0. If x x > 0, then g, ) g, ) 0, nd f x g, ) 0, whch we lredy noted re mpossble. Thus, x x g, ) g, )) = g, ) g, ). x < 0, then g, ) = x, whch mples tht Ths proves b). If g hs no domnted strtegy, then 2) s true for every, A.Ifw, ) s the sme for every, A, then better-response equvlence mples VNM-equvlence. However, Proposton 1 does not sy nythng bout whether w, ) does depend upon, A. Thus, we re nterested n when better-response equvlence mples VNMequvlence. The followng proposton provdes suffcent condton for the equvlence of better-response equvlence nd VNM-equvlence. Proposton 2. Suppose tht gmes g nd g stsfy generc propertes G1 nd G2, nd tht, for ech N nd for ny, A, there exsts sequence { k}m k=1 such tht 1 =, m =, k g k+1 for k = 1,...,m 1, nd k g k+2 for k = 1,...,m 2. Then g s better-response equvlent to g f nd only f g s VNM-equvlent to g. Note tht the bove condton concernng g s trvlly stsfed f no strtegy s domnted,.e., g s the complete relton. So, the proposton mmedtely hs the followng corollry. Corollry 3. If g nd g stsfy generc propertes G1 nd G2 nd hve no strctly domnted strteges, then g s better-response equvlent to g f nd only f g s VNMequvlent to g.

10 S. Morrs, T. U / Gmes nd Economc Behvor ) It should be emphszed tht the suffcent condton of Proposton 2 s sometmes stsfed even when there re strctly domnted strteges n the gme. For exmple, consder the followng two plyer gme, where only the row plyer s pyoffs re shown: Consder strteges of the row plyer. We hve 1 g 2, 2 g 3, 3 g 4, 1 g 3, 2 g 4 s n Fg. 2, stsfyng the condton of Proposton 2, whle strtegy 1 strctly domntes strtegy 4. To prove the proposton, we use the followng lemm. Lemm 3. Suppose tht g nd g stsfy generc property G2. For some A A,fthere exsts w, )>0 such tht g, ) g, ) = w, ) g, ) g, )) for ll, A,thenw, ) s the sme for ll, A. Proof. Wthout loss of generlty, ssume tht A 3. For dstnct,b,c A,there exst w,b ), w b,c ), w,c )>0 such tht g, ) g b, ) = w,b ) g, ) g b, ) ), g b, ) g c, ) = w b,c ) g b, ) g c, ) ), g, ) g c, ) = w,c ) g, ) g c, ) ). We frst show tht w,b ) = w b,c ) = w,c ).Snce w,b ) g, ) g b, ) ) + w b,c ) g b, ) g c, ) ) = g, ) g b, ) + g b, ) g c, ) = g, ) g c, ) = w,c ) g, ) g c, ) ) = w,c ) g, ) g b, ) ) + w,c ) g b, ) g c, ) ), Fg. 2. The grph of g.

11 270 S. Morrs, T. U / Gmes nd Economc Behvor ) we hve w,b ) w,c ) ) g, ) g b, ) ) + w b,c ) w,c ) ) g b, ) g c, ) ) = 0. By G2, g, ) g b, ) nd g b, ) g c, ) re lnerly ndependent nd thus t must be true tht w,b ) = w b,c ) = w,c ). Smlrly, for dstnct b,c,d A, w b,c ) = w c,d ) = w b,d ). Therefore, w,b ) = w c,d ) for ny,b,c,d A, whch completes the proof. We now report the proof of Proposton 2. Proof of Proposton 2. We show tht f g s better-response equvlent to g then g s VNM-equvlent to g. By G1 nd Proposton 1, f g,thereexstw, )>0 such tht g, ) g, ) = w, ) g, ) g, )). If A =2, ths completes the proof by Lemm 1. Suppose tht A 3. For, A, let { k}m k=1 be sequence such tht 1 =, m =, k g k+1 for k = 1,...,m 1, nd k g k+2 for k = 1,...,m 2. There exst x k,y k > 0 such tht g k, ) g k+1, ) = x k g k, ) g k+1, )), g k+1, ) g k+2, ) = x k+1 g k+1, ) g k+2, )), g k, ) g k+2, ) = y k g k, ) g k+2, )). By Lemm 3, x k = x k+1 = y k for ll k m 2. By lettng x k = w, ),wehve m 1 g, ) g, ) = g k, ) g k+1, )) k=1 m 1 = k=1 x k g k, ) g k+1, )) = w, ) g, ) g, )). To summrze, for ll, A, there exsts w, )>0 stsfyng the bove equton. By Lemm 3, w, ) s the sme for ll, A. By Lemm 1, g s VNM-equvlent to g, whch completes the proof Best-response equvlence Strteges nd re best-response comprble we wrte g ) f both strteges re best responses t some belef,.e., Λ,A g ) Λ,A g ). Note tht g f nd only f Λ,A g ).

12 S. Morrs, T. U / Gmes nd Economc Behvor ) Proposton 4. If gmes g nd g stsfy generc property G3, then g s best-response equvlent to g f nd only f, for ech N, ) they hve the sme best-response comprblty relton g = g ) nd b) whenever s best-response comprble to g ), there exsts w, )>0 such tht g, ) g, ) = w, ) g, ) g, )). Proof. We frst show tht ) nd b) re suffcent for the best-response equvlence of g nd g.ifλ,a g ) =,thenλ,a g ) = Λ,A g ) = becuse Λ,A g ) = mples tht g s not true nd thus ) mples tht g s not true. If Λ,A g ),then{ g }, nd we must hve Λ,A g ) = Λ, g ) = Λ, g ). 3) A { g } Clerly, 3) s true when { g }=A. To see tht 3) s true when { g } A, suppose otherwse. Then, Λ, g ) Λ, g ), A { g } nd thus there exsts / { g } such tht Λ, g ) Λ, g ). A A \ { } However, ths mples tht g, whch s contrdcton. Thus, 3) must be true. If g, then b) mples tht )) λ ) g, ) g, = w, ) g, ) g )),, nd thus λ ) Λ, g ) = Λ, g ). 4) Therefore, by ), 3), nd 4), we hve Λ,A g ) = Λ,A g ). Ths completes the proof of suffcency. To prove necessty, suppose tht g s best-response equvlent to g.snce Λ,A g ) = Λ,A g ), we hve ) Λ,A g ) Λ,A g = Λ,A g ) Λ,A g ) nd thus g = g.thsproves).

13 272 S. Morrs, T. U / Gmes nd Economc Behvor ) If g, then there exsts λ A ) such tht λ ) g, ) g, )) 0 forll A, ) λ ) g, g, )) 0 forll A \{ }. Snce λ Λ,A g ) = Λ,A g ), λ ) g, ) g )), 0. The bove mples tht, f y ) A R A s such tht )) g, ) g, 0, then y y y g, ) g, )) 0 forll A \{, }, ) g, g, )) 0 forll A \{, }, y 0 forll A, y g, ) g )), 0. By Frks Lemm, there exst x 0, γ : A R,ndδ : A R such tht, )) where wth u x g, ) g, )) γ ) δ ) = g, ) g γ ) =,,v 0nd δ ) = z δ ) wth z 0. Thus, x u g, ) g, )) +, g, ) g, )) + γ ) g, ) g, ). We show x > 0. Suppose tht x λ Λ,A \ { } g ) \Λ, g ), v g, ) g, )) = 0,.e., γ ) g, ) g, ).Let

14 S. Morrs, T. U / Gmes nd Economc Behvor ) whch exsts by g nd G3. Snce λ Λ,A \{ } g ) Λ,A \{ } g ), λ )γ ) = u λ ) g, ) g, )), + v λ ) ) g, g, )) 0., Snce λ Λ,A g ) = Λ,A g ) nd λ / Λ,A g ) = Λ,A g ), λ ) g, ) g )), < 0. Ths s contrdcton. Thus, we must hve x > 0. We hve x g, ) g, )) + γ ) g, ) g nd symmetrclly, ) x g, ) g, ) ) + γ ) g, ) g, ) where x,x > 0. Addng both, x x ) g, ) g, )) + γ We show x x = 0. Suppose tht x x > 0. Let ) ) λ Λ,A \{ } g \Λ, g Λ,A \ { } ) ) g Λ,A \{ } g. ) + γ ) 0. 5) Then, the expectton of the left-hnd sde of 5) s postve becuse ) x x )) λ ) g, ) g, > 0 nd ) λ ) γ ) + γ ) = ) u + v, λ ) g, ) g, )) + ) v + u ) λ ) g, g, )) 0.,

15 274 S. Morrs, T. U / Gmes nd Economc Behvor ) Ths s contrdcton. Symmetrclly, f x x < 0, then we hve the symmetrc contrdcton. Thus, x x = 0, nd 5) s reduced to γ ) + γ ) 0. 6) We show γ ) = γ ) = 0. Suppose tht ether γ ) 0orγ ) 0strue.Let λ,λ A ) be such tht ) ) λ Λ,A \{ } g \Λ, g Λ,A \ { } ) ) g Λ,A \{ } g, λ Λ,A \ { } ) g \Λ, g ) Λ,A \ { } ) ) g Λ,A \{ } g. Consder λ + λ )/2 A ). Then, the expectton of the left-hnd sde of 6) s postve becuse λ ) + λ 2 =, + u,, + v u, v ) γ ) + v ) ) + γ ) λ ) + λ 2 ) + u + v ) λ ) + λ 2 ) λ ) 2 + u ) λ ) Ths s contrdcton. Thus, γ ) = γ ) = 0. Summrzng the bove, we hve x g, ) g, )) = g, ) g where x > 0. Ths proves b). 2 ) g, ) g, )) ) g, g, )) g, ) g, )), ) g, ) g, )) > 0. The followng proposton nd corollry follow by exctly the sme rguments n Proposton 2 nd Corollry 3 n the prevous subsecton for better-response equvlence. Proposton 5. Suppose tht gmes g nd g stsfy generc propertes G2 nd G3, nd tht, for ech N nd for ny, A, there exsts sequence { k}m k=1 such tht

16 S. Morrs, T. U / Gmes nd Economc Behvor ) =, m =, k g k+1 for k = 1,...,m 1, k g k+2 for k = 1,...,m 2.Then g s best-response equvlent to g f nd only f g s VNM-equvlent to g. Corollry 6. If g nd g stsfy generc propertes G2 nd G3 nd g s the complete relton, then g s best-response equvlent to g f nd only f g s VNM-equvlent to g. 4. Gmes wth own-strtegy unmodlty Best-response equvlence relton s n equvlence relton. It wll be useful f, s closed form, we cn descrbe the best-response equvlence clss of gme n whch best-response equvlence s strctly weker requrement thn VNM-equvlence. Let A be lnerly ordered such tht A ={1,...,K } wth K 3. For q : A R nd w : A \{K } R ++,letq,w ) g : A R be such tht q,w ) g 1, ) = q ), 1 q,w ) g, ) = q ) + w k) g k + 1, ) g k, ) ) for 2. k=1 Let D g ) be clss of pyoff functons of plyer obtned by ths trnsformton: D g ) = { g : A R g = q },w ) g,q : A R, w : A \{K } R ++. It s strghtforwrd to see tht g D g ) f nd only f there exsts w : A \{K } R ++ such tht g + 1, ) g, ) = w ) g + 1, ) g, ) ) 7) for ll A \{K }. Note tht g D g ), g D g ) mples g D g ),ndg D g ) wth g D g ) mples g D g ). Thus, D g ) defnes n equvlence clss of pyoff functons of plyer. We wrte Dg) = { g = g ) N g D g ) for ll N }. For exmple, consder prmetrzed clss of gmes {gx, y)} x,y) R 2 dscussed n ++ Secton 2. We hve tht {gx, y)} x,y) R 2 Dg1, 1)). To see ths, we wrte gx, y) = ++ g x,y)) {1,2}. Then, for ny x, y) R 2 ++ nd j, g 1, j x,y) = q j ), g 2, j x,y) = q j ) + x g 2, j 1, 1) g 1, j 1, 1) ), g 3, j x,y) = q j ) + x g 2, j 1, 1) g 1, j 1, 1) ) + y g 3, j 1, 1) g 2, j 1, 1) ) where q : {1, 2, 3} R s such tht q 1) = x, q 2) = x, ndq 3) = 2x. Remember tht, for ny x, y) R 2 ++, gx, y) s best-response equvlent to g1, 1). It s esy to see tht every gme n Dg1, 1)) s VNM-equvlent to gx, y) for some x, y) R Thus, every gme n Dg1, 1)) s best-response equvlent to g1, 1).

17 276 S. Morrs, T. U / Gmes nd Economc Behvor ) Ths observton leds us to the queston when every gme n Dg) s best-response equvlentto g. We provde necessry nd suffcent condton for t. We sy tht g s own-strtegy unmodl f, for ll λ A ), there exsts k A such tht, λ ) g, ) g 1, ) ) 0 f k nd k > 1, λ ) g, ) g + 1, ) ) 8) 0 f k nd k <K. Note tht f g s own-strtegy unmodl, then 8) s true f nd only f λ Λ k,a g ). Clerly, by 7), g s own-strtegy unmodl f nd only f g D g ) s own-strtegy unmodl. We sy tht g s own-strtegy concve f g, ) : A R s concve,.e., g + 1, ) g, ) s decresng n for ll A. Lemm 4. Suppose tht g + 1, ) g, ) for ll A \{K } nd A, nd tht there s no wekly domnted strtegy. Then, g s own-strtegy unmodl f nd only f there exsts g D g ) such tht g s own-strtegy concve. Proof. Suppose tht g D g ) s own-strtegy concve. Then, g + 1, ) g, ) s decresng n for ll A. Thus, λ ) g + 1, ) g, )) s lso decresng n for ll λ A ). Ths mmedtely mples tht g D g ) s own-strtegy unmodl. Snce λ ) g + 1, ) g, ) ) = 1 w ) λ ) g + 1, ) g, ) ), g s lso own-strtegy unmodl. Suppose tht g s own-strtegy unmodl. We prove the exstence of n own-strtegy concve pyoff functon g = q,w ) g by constructon. Lter, we wll show tht there exsts C k > 0 such tht g k + 1, ) g k, ) C k g k + 2, ) g k + 1, ) ). 9) For C k stsfyng 9), we let w : A R ++ be such tht w 1) = 1ndw ) = 1 k=1 C k for 2, nd q : A R be such tht q ) = 0forll A.Snce g + 1, ) g, ) = w ) g + 1, ) g, ), we hve g k + 1, ) g k, ) = w k) g k + 1, ) g k, ) ), g k + 2, ) g k + 1, ) = C k w k) g k + 2, ) g k + 1, ) ).

18 S. Morrs, T. U / Gmes nd Economc Behvor ) By ths nd 9), we hve g k + 1, ) g k, ) g k + 2, ) g k + 1, ), whch mples tht g s own-strtegy concve. We prove the exstence of C k stsfyng 9) by Frks Lemm. Before dong t, we must frst observe tht f λ ) g k + 1, ) g k, ) ) = 0 10) then λ ) g k + 2, ) g k + 1, ) ) 0. To see ths, suppose otherwse. Then, there exsts λ A ) stsfyng both 10) nd λ ) g k + 2, ) g k + 1, ) ) > 0. Snce g k + 1, ) g k, ) 0forll A, 10) mples tht there exst, A such tht 0 <λ )<1 wth g k + 1, ) g k, )>0nd 0 <λ )<1 wth g k + 1, ) g k, )<0. Let ε>0 be suffcently smll. More precsely, let ε>0 be such tht { } ) ) ε<mn λ, 1 λ, A λ )g k + 2, ) g k + 1, )). 2 mx A g k + 2, ) g k + 1, ) Let λ A ) be such tht λ ) ε f = λ, ) = λ ) + ε f =, λ ) otherwse. Then, we hve λ ) g k + 1, ) g k, ) ) = λ ) g k + 1, ) g k, ) ) + ε ) )) ) )) g k + 1, g k, ε g k + 1, g k, = ε ) ) )) g k + 1, g k, )) ε g k + 1, g k, < 0, λ ) g k + 2, ) g k + 1, ) ) = λ ) g k + 2, ) g k + 1, ) ) + ε ) )) ) )) g k + 2, g k + 1, ε g k + 2, g k + 1,

19 278 S. Morrs, T. U / Gmes nd Economc Behvor ) λ ) g k + 2, ) g k + 1, ) ) 2ε mx g k + 2, ) g k + 1, ) > 0, whch contrdcts to the ssumpton tht g s own-strtegy unmodl. Now, we know tht, f g s own-strtegy unmodl nd stsfes the ssumptons, then t must be true tht f λ ) g k + 1, ) g k, ) ) 0, then λ ) g k + 2, ) g k + 1, ) ) 0. Ths mples tht f y ) A R A s such tht g k + 1, ) g k, ) ) 0, then y y 0 forll A, y g k + 2, ) g k + 1, ) ) 0. By Frks Lemm, there exst x k 0ndz 0for A such tht x k g k + 1, ) g k, ) ) z δ ) = g k + 2, ) g k + 1, ). Thus, x k g k + 1, ) g k, ) ) g k + 2, ) g k + 1, ). 11) If x k = 0, then g k + 2, ) g k + 1, ) 0. However, ths s mpossble snce there s no wekly domnted strtegy. Thus, x k > 0. By lettng C k = 1/x k, 11) mples 9). Consder gn {gx, y)} x,y) R 2 ++ Dg1, 1)). In generl, g x,y) s not lwys own-strtegy concve. However, g 1, 1) s own-strtegy concve. Thus, Lemm 4 sys tht g x,y) s own-strtegy unmodl. We clm tht, generclly, Dg) s best-response equvlence clss f nd only f g s own-strtegy unmodl for ll N. Proposton 7. Suppose tht g hs no domnted strtegy. Every gme n Dg) s bestresponse equvlent to g f nd only f g s own-strtegy unmodl for ll N. Ifg s own-strtegy unmodl for ll N nd g stsfes generc property G3, then every gme best-response equvlent to g nd stsfyng G3 s n Dg).

20 S. Morrs, T. U / Gmes nd Economc Behvor ) Proof. Suppose tht g s own-strtegy unmodl for ll N. We show tht f g Dg) then g s best-response equvlent to g.letλ Λ,A g ). Then, 8) mples tht λ ) g, ) g 1, ) ) 0 f nd > 1, λ ) g, ) g + 1, ) ) 0 f nd <K 12). By 7), ths s true f nd only f λ ) g, ) g 1, ) ) 0 f nd > 1, λ ) g, ) g + 1, ) ) 0 f nd <K. Thus, λ Λ,A g ).Conversely,letλ Λ,A g ).Snceg s own-strtegy unmodl, we hve 13), whch s true f nd only f 12) s true. Thus, λ Λ,A g ). Therefore, Λ,A g ) = Λ,A g ) nd thus g s best-response equvlent to g. Conversely, suppose tht every gme n Dg) s best-response equvlent to g.weshow tht g s own-strtegy unmodl for ll N. Seekng contrdcton, suppose otherwse. Then, there exst, ã A nd λ Λ,A g ) such tht ether of the followng s true: < ã nd λ ) g ã, ) g ã 1, ) ) > 0, 14) > ã nd λ ) g ã, ) g ã + 1, ) ) > 0. 15) When 14) s true, let g = q,w ) g D g ) be such tht q ) = 0nd { w ) = L f =ã 1, 1 otherwse. Then, we hve λ ) g ã, ) g, ) ) = λ ) g ã, ) g ã 1, ) ) + λ ) g ã 1, ) g, ) ) = L λ ) g ã, ) g ã 1, ) ) + λ ) g ã 1, ) g, ) ). 13)

21 280 S. Morrs, T. U / Gmes nd Economc Behvor ) By choosng very lrge L>0, we hve λ ) g ã, ) g, ) ) > 0 nd thus Λ,A g ) Λ,A g ). When 15) s true, we lso hve Λ,A g ) Λ,A g ) by the smlr rgument. Ths mples tht some gme n Dg) s not best-response equvlent to g, whch completes the proof of the frst hlf of the proposton. We prove the lst hlf of the proposton. Suppose tht g s own-strtegy unmodl for ll N nd tht g stsfes generc property G3. Let g be best-response equvlent to g nd stsfy G3. We show g Dg). We frst observe tht g +1forll A \{K }. To see ths, let λ k Λ k, A g ) for k A, whch exsts snce g hs no domnted strtegy. Note tht f λ = λ k or λ = λ k+1 then λ )g k, ) λ )g, ) for ll k, 16) λ )g k + 1, ) λ )g, ) for ll k + 1. Let t [0, 1] nd λ k,t = tλ k + 1 t)λk+1 A ) be such tht )g k, ) = )g k + 1, ). 17) λ k,t λ k,t Then, 16) mples tht )g k, ) λ k,t λ k,t λ k,t )g k + 1, ) )g, ) for ll k, λ k,t )g, ) for ll k + 1. By 17), we hve λ k,t Λ k, A g ) Λ k + 1,A g ). Thsmplestht g + 1 for ll A \{K }. Snce g nd g stsfy G3 nd re best-response equvlent, we cn use Proposton 4, whch sys tht there exsts w : A \{K } R ++ such tht g + 1, ) g, ) = w ) g + 1, ) g, ) ). Ths mples tht g D g ) nd thus g Dg). A weker, but smlr clm s true for gmes such tht strtegy sets re ntervls of rel numbers nd pyoff functons re dfferentble, whch hs couple of pplctons. In the remnder of ths secton, we dscuss ths ssue. Abusng notton, we gve defnton of best-response equvlence for clss of gmes wth contnuum of ctons. Let A be closed ntervl of R for ll N. Assume tht g : A R s bounded nd contnuously dfferentble. Let A ) be the set of ll probblty mesures over A nd Λ,X g ) be such tht

22 S. Morrs, T. U / Gmes nd Economc Behvor ) Λ,X g ) { = λ A ) g, ) g, )) dλ ) 0forll X }. A The defnton of best-response equvlence s the sme s tht for fnte gmes: we sy tht g s best-response equvlent to g f, for ech N, Λ,A g ) = Λ,A g ) for ll A. We sy tht g s own-strtegy unmodl f, for ny λ A ), there exsts x such tht g, ) dλ ) 0 f x nd x > mn A, A 18) g, ) dλ ) 0 f x nd x < mx A. A Note tht f g s own-strtegy unmodl, then 18) s true f nd only f λ Λ x,a g ). Snce g, ) g, ) dλ ) = dλ ), A A g s own-strtegy unmodl f g s own-strtegy concve,.e., g, )/ s decresng n for ll A. For mesurble functons q : A R nd w : A R ++,letq,w ) g : A R be such tht, for A nd A, Let q,w ) g, ) = q ) + x w x) g x, ) x D g ) = { g : A R g = q },w ) g,q : A R, w : A R ++, Dg) = { g = g ) N g D g ) }. Proposton 8. Suppose tht g s own-strtegy unmodl for ll N. Then, every gme n Dg) s best-response equvlent to g. Proof. Let g Dg). Snceg s own-strtegy unmodl, for ll λ A ), there exsts A such tht g, ) dλ ) 0 f nd > mn A, A 19) g, ) dλ ) 0 f nd < mx A. A dx.

23 282 S. Morrs, T. U / Gmes nd Economc Behvor ) Snce g, ) = w ) g, ), 19) s true f nd only f g, ) dλ ) 0 f nd > mn A, A g, ) dλ ) 0 f nd < mx A. A Thus, g s lso own-strtegy unmodl. Snce 19) s true f nd only f λ Λ,A g ) nd 20) s true f nd only f λ Λ,A g ),wemusthveλ,a g ) = Λ,A g ), whch completes the proof. Ths proposton hs useful pplcton concernng the unqueness of correlted equlbr. Neymn 1997) showed tht f g hs contnuously dfferentble nd strctly concve potentl functon, 9 then the potentl mxmzer s the unque correlted equlbrum of g. The set of correlted equlbr s the sme for two gmes f the two gmes re best-response equvlent. Thus, we clm the followng. Corollry 9. Suppose tht g hs contnuously dfferentble nd strctly concve potentl functon f. Then, the potentl mxmzer s the unque correlted equlbrum of every gme n Dg). Note tht gme n Dg) s not necessrly potentl gme nd pyoff functons re not necessrly concve. 20) 5. Mxed extensons of equvlence We hve focused on plyers preferences over pure strteges, gven nondegenerte conjectures bout ther opponents behvor. But we could sk the sme queston n the mxed strtegy extenson of the orgnl gme; equvlently, we could look t plyers preferences over mxed strteges. 10 The nturl queston s whether or not our dscusson so fr must be modfed by the mxed extenson of equvlence. For N,let A ) denote the set of ll mxed strteges of plyer. Abusng notton, we wrte g p, ) = A p )g, ) for p A ). By the mxed extenson of Λ, we cn nturlly defne Λ p,x g ) for p A ) nd X A ): Λ p,x g ) 9 The defnton of potentl functons of ths clss of gmes s the sme s those of fnte gmes. 10 The ssocte edtor suggested the observtons n ths secton.

24 { = λ A ) S. Morrs, T. U / Gmes nd Economc Behvor ) } )) λ ) g p, ) g p, 0forllp X. We use the sme rule Λ p,p g ) = Λ p, {p } g ) s before. We consder the followng equvlence reltons of gmes. Defnton 7. Agmeg s mxed better-response equvlent to g f, for ech N, Λ p,p g ) = Λ p,p ) g for ll p,p A ). Defnton 8. Agmeg s mxed best-response equvlent to g f, for ech N, ) Λ p, A ) g = Λ p, A ) g ) for ll p A ). Note tht VNM-equvlence s suffcent for both mxed better-response equvlence nd mxed best-response equvlence. Note lso tht mxed better-response equvlence s suffcent for better-response equvlence, nd tht mxed best-response equvlence s suffcent for best-response equvlence. It s esy to see tht mxed best-response equvlence s not only suffcent but lso necessry for best-response equvlence. Lemm 5. A gme g s mxed best-response equvlent to g f nd only f g s best-response equvlent to g. Proof. Note tht λ Λ p, A ) g ) ) p rg mx λ )g p, p A ) ) p )>0 mples rg mx λ )g, A p )>0 mples λ Λ,A g ). Thus, f Λ,A g ) = Λ,A g ) for ll A,thenΛ p, A ) g ) = Λ p, A ) g ) for ll p A ). Ths completes the proof. Ths lemm mples tht the chrcterzton of mxed best-response equvlence s reduced to tht of best-response equvlence. On the other hnd, mxed better-response equvlence s strctly stronger requrement thn better-response equvlence. Consder two plyer, three strtegy, symmetrc pyoff gmes g nd g, where ech plyer s pyoffs re gven by the followng pyoff mtrces

25 284 S. Morrs, T. U / Gmes nd Economc Behvor ) where the plyer s own strteges re represented by rows nd hs opponent s strteges re represented by columns): g g x 2x 2x y y y We ssume tht x,y > 0ndx/2 <y<2x. Then, 1 g 2, 2 g 3, 3 g 1, nd g = g. We lso hve xg 1, ) g 2, )) = g 1, ) g 2, ) nd yg 2, ) g 3, )) = g 2, ) g 3, ). Thus, by Proposton 1, g s better-response equvlent to g. However, we cn show tht g s mxed better-response equvlent to g only f x = y. To see ths, suppose tht row plyer beleves tht the column plyer never chooses 1: row plyer hs belef λ wth λ 1) = 0. Consder row plyer s mxed strtegy p such tht p 1) = p nd p 2) = 1 p. Ing, he prefers strtegy p to strtegy 3 f nd only f 2p 1,.e., p 1/2. In g, he prefers strtegy p to strtegy 3 f nd only f 2xp y,.e., p y/2x. In order for g to be mxed better-response equvlent to g,tmustbetruetht 1/2 = y/2x,.e., x = y. In ths cse, g s VNM-equvlent to g ndthusmxedbetterresponse equvlent to g. In the bove exmple, the relton g genertes connected grph snce 1 g 2 nd 2 g 3. Thus, under the connectedness of g, better-response equvlence does not necessrly mply VNM-equvlence, but mxed better-response equvlence my mply VNM-equvlence. The nturl queston s whether ths s true. Remember tht Proposton 2 provdes condton to ensure the equvlence of better-response equvlence nd VNM-equvlence. The condton ncludes the connectedness of g.but the connectedness s not suffcent s demonstrted by the bove exmple. In contrst, the followng proposton sserts tht the connectedness of g ensures the equvlence of mxed better-response equvlence nd VNM-equvlence. Proposton 10. Suppose tht gmes g nd g stsfy generc propertes G1 nd G2, nd tht, for ech N, g genertes connected grph on A.Theng s mxed betterresponse equvlent to g f nd only f g s VNM-equvlent to g. To prove the proposton, we use the followng lemm. Lemm 6. Suppose tht g nd g stsfy generc propertes G1 nd G2, nd tht g s mxed better-response equvlent to g. For dstnct,b,c A,f g b nd b g c,then there exsts w > 0 such tht g, ) g b, ) = w g, ) g b, ) ), g b, ) g c, ) = w g b, ) g c, ) ). Proof. By G1 nd Proposton 1, there exst w,b ), w b,c )>0 such tht

26 S. Morrs, T. U / Gmes nd Economc Behvor ) g, ) g b, ) = w,b ) g, ) g b, ) ), g b, ) g c, ) = w b,c ) g b, ) g c, ) ). We show tht w,b ) = w b,c ). Ether g c, g c,orc g s true. If g c, there exsts w,c )>0such tht g, ) g c, ) = w,c ) g, ) g c, ) ) by Proposton 1. Thus, by Lemm 3, w,b ) = w b,c ) = w,c ). Suppose tht g c. Note tht g = g by Proposton 1. Let λ A ) be such tht λ ) g b, ) g c, ) ) < 0, 21) whch exsts snce b g c nd G1. The relton g c mples tht λ ) g, ) g c, ) ) > 0. 22) By the weghted verge of 21) nd 22), we cn choose p A ) such tht p ) = p, p b ) = 1 p, nd λ ) g p, ) g c, ) ) = 0. 23) Mxed better-response equvlence mples tht λ ) g p, ) g c, ) ) = 0. 24) Now clculte g p, ) g c, ) = pg, ) + 1 p)g b, ) g c, ) = p g, ) g b, ) ) + g b, ) g c, ) = w,b )p g, ) g b, ) ) + w b,c ) g b, ) g c, ) ) = w,b ) pg, ) + 1 p)g b, ) g c, ) ) + w b,c ) w,b ) ) g b, ) g c, ) ) = w,b ) g p, ) g c, ) ) + w b,c ) w,b ) ) g b, ) g c, ) ). By the expecttons wth respect to λ for both sdes of the equton, nd by 23) nd 24), we hve w b,c ) w,b ) ) λ ) g b, ) g c, ) ) = 0.

27 286 S. Morrs, T. U / Gmes nd Economc Behvor ) By 21), we must hve w,b ) = w b,c ). Smlrly, f c g,wemusthve w,b ) = w b,c ). Ths completes the proof. We now report the proof of Proposton 10. Proof of Proposton 10. We show tht f g s mxed better-response equvlent to g then g s VNM-equvlent to g. By G1 nd Proposton 1, f g, there exsts w, )>0 such tht g, ) g, ) = w, ) g, ) g, )). If A =2, ths completes the proof by Lemm 1. Suppose tht A 3. For, A, let { k}m k=1 be sequence such tht 1 =, m =, k g k+1 for k = 1,...,m 1, whch exsts by the connectedness of g. There exsts x k > 0 such tht g k, ) g k+1, ) = x k g k, ) g k+1, )). By Lemm 6, x k = x k+1 for ll k m 1. By lettng x k = w, ),wehve m 1 g, ) g, ) = g k, ) g k+1, )) k=1 m 1 = k=1 x k g k, ) g k+1, )) = w, ) g, ) g, )). To summrze, for ll, A, there exsts w, )>0 stsfyng the bove equton. By Lemm 3, w, ) s the sme for ll, A. By Lemm 1, g s VNM-equvlent to g, whch completes the proof. Acknowledgments We re very grteful for vluble nput from Lrry Blume, George Mlth nd Phlp Reny. U cknowledges fnncl support by the Mnstry of Educton, Scence, Sports nd Culture, Grnt-n-Ad for Scentfc Reserch. References Anderson, S.P., Jcob, K., Holt, C.A., Mnmum-effort coordnton gmes: stochstc potentl nd logt equlbrum. Gmes Econ. Behv. 34, Blume, L., The sttstcl mechncs of strtegc ntercton. Gmes Econ. Behv. 5, Brock, W., Durluf, S., Dscrete choce wth socl nterctons. Rev. Econ. Stud. 68, Dubey, P., Hmnko, O., Zpechelnyuk, A., Strtegc substtutes nd potentl gmes. Mmeo. SUNY t Stony Brook. Hrrt-Urruty, J.-B., Lemréchl, C., Fundmentls of Convex Anlyss. Sprnger-Verlg, New York. Mskn, E., Trole, J., Mrkov perfect equlbrum. J. Econ. Theory 100,

28 S. Morrs, T. U / Gmes nd Economc Behvor ) Mertens, J.-F., Ordnlty n non-coopertve gmes. Mmeo. DP8728, CORE, Unversté Ctholque de Louvn. Monderer, D., Shpley, L.S., Fcttous ply property for gmes wth dentcl nterests. J. Econ. Theory 68, Monderer, D., Shpley, L.S., 1996b. Potentl gmes. Gmes Econ. Behv. 14, Morrs, S., Potentl methods n ntercton gmes. Mmeo. Yle Unversty. Avlble from econ.yle.edu/~sm326/pot-method.pdf. Morrs, S., U, T., Generlzed potentls nd robust sets of equlbr. Mmeo. Yle Unversty. Avlble from Neymn, A., Correlted equlbrum nd potentl gmes. Int. J. Gme Theory 26, Rockfellr, R.T., Convex Anlyss. Prnceton Unv. Press, Prnceton, NJ. Rosenthl, R.W., A clss of gmes possessng pure strtegy equlbr. Int. J. Gme Theory 2, Sel, A., Lernng processes n gmes. MSc thess. The Technon, Hf, Isrel. [In Hebrew.] Sel, A., Fcttous ply n one-gnst-ll mult-plyer gmes. Econ. Theory 14, U, T., A Shpley vlue representton of potentl gmes. Gmes Econ. Behv. 31, U, T., Robust equlbr of potentl gmes. Econometrc 69, U, T., Quntl response equlbr nd stochstc best response dynmcs. Mmeo. Yokohm Ntonl Unversty. Avlble from Voorneveld, M., Best-response potentl gmes. Econ. Letters 66,

COWLES FOUNDATION FOR RESEARCH IN ECONOMICS YALE UNIVERSITY

COWLES FOUNDATION FOR RESEARCH IN ECONOMICS YALE UNIVERSITY Best Response Equvalence by Stephen Morrs and Takash U July 2002 COWLES FOUNDATION DISCUSSION PAPER NO. 1377 COWLES FOUNDATION FOR RESEARCH IN ECONOMICS YALE UNIVERSITY Box 208281 New Haven, Connectcut

More information

Two Coefficients of the Dyson Product

Two Coefficients of the Dyson Product Two Coeffcents of the Dyson Product rxv:07.460v mth.co 7 Nov 007 Lun Lv, Guoce Xn, nd Yue Zhou 3,,3 Center for Combntorcs, LPMC TJKLC Nnk Unversty, Tnjn 30007, P.R. Chn lvlun@cfc.nnk.edu.cn gn@nnk.edu.cn

More information

Rank One Update And the Google Matrix by Al Bernstein Signal Science, LLC

Rank One Update And the Google Matrix by Al Bernstein Signal Science, LLC Introducton Rnk One Updte And the Google Mtrx y Al Bernsten Sgnl Scence, LLC www.sgnlscence.net here re two dfferent wys to perform mtrx multplctons. he frst uses dot product formulton nd the second uses

More information

UNIVERSITY OF IOANNINA DEPARTMENT OF ECONOMICS. M.Sc. in Economics MICROECONOMIC THEORY I. Problem Set II

UNIVERSITY OF IOANNINA DEPARTMENT OF ECONOMICS. M.Sc. in Economics MICROECONOMIC THEORY I. Problem Set II Mcroeconomc Theory I UNIVERSITY OF IOANNINA DEPARTMENT OF ECONOMICS MSc n Economcs MICROECONOMIC THEORY I Techng: A Lptns (Note: The number of ndctes exercse s dffculty level) ()True or flse? If V( y )

More information

Online Appendix to. Mandating Behavioral Conformity in Social Groups with Conformist Members

Online Appendix to. Mandating Behavioral Conformity in Social Groups with Conformist Members Onlne Appendx to Mndtng Behvorl Conformty n Socl Groups wth Conformst Members Peter Grzl Andrze Bnk (Correspondng uthor) Deprtment of Economcs, The Wllms School, Wshngton nd Lee Unversty, Lexngton, 4450

More information

Statistics and Probability Letters

Statistics and Probability Letters Sttstcs nd Probblty Letters 79 (2009) 105 111 Contents lsts vlble t ScenceDrect Sttstcs nd Probblty Letters journl homepge: www.elsever.com/locte/stpro Lmtng behvour of movng verge processes under ϕ-mxng

More information

The Number of Rows which Equal Certain Row

The Number of Rows which Equal Certain Row Interntonl Journl of Algebr, Vol 5, 011, no 30, 1481-1488 he Number of Rows whch Equl Certn Row Ahmd Hbl Deprtment of mthemtcs Fcult of Scences Dmscus unverst Dmscus, Sr hblhmd1@gmlcom Abstrct Let be X

More information

Demand. Demand and Comparative Statics. Graphically. Marshallian Demand. ECON 370: Microeconomic Theory Summer 2004 Rice University Stanley Gilbert

Demand. Demand and Comparative Statics. Graphically. Marshallian Demand. ECON 370: Microeconomic Theory Summer 2004 Rice University Stanley Gilbert Demnd Demnd nd Comrtve Sttcs ECON 370: Mcroeconomc Theory Summer 004 Rce Unversty Stnley Glbert Usng the tools we hve develoed u to ths ont, we cn now determne demnd for n ndvdul consumer We seek demnd

More information

Jean Fernand Nguema LAMETA UFR Sciences Economiques Montpellier. Abstract

Jean Fernand Nguema LAMETA UFR Sciences Economiques Montpellier. Abstract Stochstc domnnce on optml portfolo wth one rsk less nd two rsky ssets Jen Fernnd Nguem LAMETA UFR Scences Economques Montpeller Abstrct The pper provdes restrctons on the nvestor's utlty functon whch re

More information

The Schur-Cohn Algorithm

The Schur-Cohn Algorithm Modelng, Estmton nd Otml Flterng n Sgnl Processng Mohmed Njm Coyrght 8, ISTE Ltd. Aendx F The Schur-Cohn Algorthm In ths endx, our m s to resent the Schur-Cohn lgorthm [] whch s often used s crteron for

More information

Dennis Bricker, 2001 Dept of Industrial Engineering The University of Iowa. MDP: Taxi page 1

Dennis Bricker, 2001 Dept of Industrial Engineering The University of Iowa. MDP: Taxi page 1 Denns Brcker, 2001 Dept of Industrl Engneerng The Unversty of Iow MDP: Tx pge 1 A tx serves three djcent towns: A, B, nd C. Ech tme the tx dschrges pssenger, the drver must choose from three possble ctons:

More information

The Regulated and Riemann Integrals

The Regulated and Riemann Integrals Chpter 1 The Regulted nd Riemnn Integrls 1.1 Introduction We will consider severl different pproches to defining the definite integrl f(x) dx of function f(x). These definitions will ll ssign the sme vlue

More information

A Family of Multivariate Abel Series Distributions. of Order k

A Family of Multivariate Abel Series Distributions. of Order k Appled Mthemtcl Scences, Vol. 2, 2008, no. 45, 2239-2246 A Fmly of Multvrte Abel Seres Dstrbutons of Order k Rupk Gupt & Kshore K. Ds 2 Fculty of Scence & Technology, The Icf Unversty, Agrtl, Trpur, Ind

More information

Applied Statistics Qualifier Examination

Applied Statistics Qualifier Examination Appled Sttstcs Qulfer Exmnton Qul_june_8 Fll 8 Instructons: () The exmnton contns 4 Questons. You re to nswer 3 out of 4 of them. () You my use ny books nd clss notes tht you mght fnd helpful n solvng

More information

Remember: Project Proposals are due April 11.

Remember: Project Proposals are due April 11. Bonformtcs ecture Notes Announcements Remember: Project Proposls re due Aprl. Clss 22 Aprl 4, 2002 A. Hdden Mrov Models. Defntons Emple - Consder the emple we tled bout n clss lst tme wth the cons. However,

More information

Lecture 4: Piecewise Cubic Interpolation

Lecture 4: Piecewise Cubic Interpolation Lecture notes on Vrtonl nd Approxmte Methods n Appled Mthemtcs - A Perce UBC Lecture 4: Pecewse Cubc Interpolton Compled 6 August 7 In ths lecture we consder pecewse cubc nterpolton n whch cubc polynoml

More information

Katholieke Universiteit Leuven Department of Computer Science

Katholieke Universiteit Leuven Department of Computer Science Updte Rules for Weghted Non-negtve FH*G Fctorzton Peter Peers Phlp Dutré Report CW 440, Aprl 006 Ktholeke Unverstet Leuven Deprtment of Computer Scence Celestjnenln 00A B-3001 Heverlee (Belgum) Updte Rules

More information

GAUSS ELIMINATION. Consider the following system of algebraic linear equations

GAUSS ELIMINATION. Consider the following system of algebraic linear equations Numercl Anlyss for Engneers Germn Jordnn Unversty GAUSS ELIMINATION Consder the followng system of lgebrc lner equtons To solve the bove system usng clsscl methods, equton () s subtrcted from equton ()

More information

CHOVER-TYPE LAWS OF THE ITERATED LOGARITHM FOR WEIGHTED SUMS OF ρ -MIXING SEQUENCES

CHOVER-TYPE LAWS OF THE ITERATED LOGARITHM FOR WEIGHTED SUMS OF ρ -MIXING SEQUENCES CHOVER-TYPE LAWS OF THE ITERATED LOGARITHM FOR WEIGHTED SUMS OF ρ -MIXING SEQUENCES GUANG-HUI CAI Receved 24 September 2004; Revsed 3 My 2005; Accepted 3 My 2005 To derve Bum-Ktz-type result, we estblsh

More information

Principle Component Analysis

Principle Component Analysis Prncple Component Anlyss Jng Go SUNY Bufflo Why Dmensonlty Reducton? We hve too mny dmensons o reson bout or obtn nsghts from o vsulze oo much nose n the dt Need to reduce them to smller set of fctors

More information

Research Article On the Upper Bounds of Eigenvalues for a Class of Systems of Ordinary Differential Equations with Higher Order

Research Article On the Upper Bounds of Eigenvalues for a Class of Systems of Ordinary Differential Equations with Higher Order Hndw Publshng Corporton Interntonl Journl of Dfferentl Equtons Volume 0, Artcle ID 7703, pges do:055/0/7703 Reserch Artcle On the Upper Bounds of Egenvlues for Clss of Systems of Ordnry Dfferentl Equtons

More information

THE COMBINED SHEPARD ABEL GONCHAROV UNIVARIATE OPERATOR

THE COMBINED SHEPARD ABEL GONCHAROV UNIVARIATE OPERATOR REVUE D ANALYSE NUMÉRIQUE ET DE THÉORIE DE L APPROXIMATION Tome 32, N o 1, 2003, pp 11 20 THE COMBINED SHEPARD ABEL GONCHAROV UNIVARIATE OPERATOR TEODORA CĂTINAŞ Abstrct We extend the Sheprd opertor by

More information

Quantum Codes from Generalized Reed-Solomon Codes and Matrix-Product Codes

Quantum Codes from Generalized Reed-Solomon Codes and Matrix-Product Codes 1 Quntum Codes from Generlzed Reed-Solomon Codes nd Mtrx-Product Codes To Zhng nd Gennn Ge Abstrct rxv:1508.00978v1 [cs.it] 5 Aug 015 One of the centrl tsks n quntum error-correcton s to construct quntum

More information

Quadratic Forms. Quadratic Forms

Quadratic Forms. Quadratic Forms Qudrtic Forms Recll the Simon & Blume excerpt from n erlier lecture which sid tht the min tsk of clculus is to pproximte nonliner functions with liner functions. It s ctully more ccurte to sy tht we pproximte

More information

A Robust Folk Theorem for the Prisoner s Dilemma

A Robust Folk Theorem for the Prisoner s Dilemma A Robust Folk Theorem for the Prsoner s Dlemm Jeffrey C. Ely Juuso Välmäk December 23, 1999 Abstrct We prove the folk theorem for the Prsoner s dlemm usng strteges tht re robust to prvte montorng. From

More information

Chapter Newton-Raphson Method of Solving a Nonlinear Equation

Chapter Newton-Raphson Method of Solving a Nonlinear Equation Chpter.4 Newton-Rphson Method of Solvng Nonlner Equton After redng ths chpter, you should be ble to:. derve the Newton-Rphson method formul,. develop the lgorthm of the Newton-Rphson method,. use the Newton-Rphson

More information

Perfect Competition and the Nash Bargaining Solution

Perfect Competition and the Nash Bargaining Solution Perfect Competton and the Nash Barganng Soluton Renhard John Department of Economcs Unversty of Bonn Adenauerallee 24-42 53113 Bonn, Germany emal: rohn@un-bonn.de May 2005 Abstract For a lnear exchange

More information

DCDM BUSINESS SCHOOL NUMERICAL METHODS (COS 233-8) Solutions to Assignment 3. x f(x)

DCDM BUSINESS SCHOOL NUMERICAL METHODS (COS 233-8) Solutions to Assignment 3. x f(x) DCDM BUSINESS SCHOOL NUMEICAL METHODS (COS -8) Solutons to Assgnment Queston Consder the followng dt: 5 f() 8 7 5 () Set up dfference tble through fourth dfferences. (b) Wht s the mnmum degree tht n nterpoltng

More information

Math 497C Sep 17, Curves and Surfaces Fall 2004, PSU

Math 497C Sep 17, Curves and Surfaces Fall 2004, PSU Mth 497C Sep 17, 004 1 Curves nd Surfces Fll 004, PSU Lecture Notes 3 1.8 The generl defnton of curvture; Fox-Mlnor s Theorem Let α: [, b] R n be curve nd P = {t 0,...,t n } be prtton of [, b], then the

More information

The Multiple Classical Linear Regression Model (CLRM): Specification and Assumptions. 1. Introduction

The Multiple Classical Linear Regression Model (CLRM): Specification and Assumptions. 1. Introduction ECONOMICS 5* -- NOTE (Summary) ECON 5* -- NOTE The Multple Classcal Lnear Regresson Model (CLRM): Specfcaton and Assumptons. Introducton CLRM stands for the Classcal Lnear Regresson Model. The CLRM s also

More information

Proof that if Voting is Perfect in One Dimension, then the First. Eigenvector Extracted from the Double-Centered Transformed

Proof that if Voting is Perfect in One Dimension, then the First. Eigenvector Extracted from the Double-Centered Transformed Proof tht f Votng s Perfect n One Dmenson, then the Frst Egenvector Extrcted from the Doule-Centered Trnsformed Agreement Score Mtrx hs the Sme Rn Orderng s the True Dt Keth T Poole Unversty of Houston

More information

Effects of polarization on the reflected wave

Effects of polarization on the reflected wave Lecture Notes. L Ros PPLIED OPTICS Effects of polrzton on the reflected wve Ref: The Feynmn Lectures on Physcs, Vol-I, Secton 33-6 Plne of ncdence Z Plne of nterfce Fg. 1 Y Y r 1 Glss r 1 Glss Fg. Reflecton

More information

Attribute reduction theory and approach to concept lattice

Attribute reduction theory and approach to concept lattice Scence n Chn Ser F Informton Scences 2005 Vol48 No6 713 726 713 Attrbute reducton theory nd pproch to concept lttce ZHANG Wenxu 1, WEI Lng 1,2 & QI Jnun 3 1 Insttute for Informton nd System Scences, Fculty

More information

UNIFORM CONVERGENCE. Contents 1. Uniform Convergence 1 2. Properties of uniform convergence 3

UNIFORM CONVERGENCE. Contents 1. Uniform Convergence 1 2. Properties of uniform convergence 3 UNIFORM CONVERGENCE Contents 1. Uniform Convergence 1 2. Properties of uniform convergence 3 Suppose f n : Ω R or f n : Ω C is sequence of rel or complex functions, nd f n f s n in some sense. Furthermore,

More information

Mixed Type Duality for Multiobjective Variational Problems

Mixed Type Duality for Multiobjective Variational Problems Ž. ournl of Mthemtcl Anlyss nd Applctons 252, 571 586 2000 do:10.1006 m.2000.7000, vlle onlne t http: www.delrry.com on Mxed Type Dulty for Multoectve Vrtonl Prolems R. N. Mukheree nd Ch. Purnchndr Ro

More information

4. Eccentric axial loading, cross-section core

4. Eccentric axial loading, cross-section core . Eccentrc xl lodng, cross-secton core Introducton We re strtng to consder more generl cse when the xl force nd bxl bendng ct smultneousl n the cross-secton of the br. B vrtue of Snt-Vennt s prncple we

More information

Review of linear algebra. Nuno Vasconcelos UCSD

Review of linear algebra. Nuno Vasconcelos UCSD Revew of lner lgebr Nuno Vsconcelos UCSD Vector spces Defnton: vector spce s set H where ddton nd sclr multplcton re defned nd stsf: ) +( + ) (+ )+ 5) λ H 2) + + H 6) 3) H, + 7) λ(λ ) (λλ ) 4) H, - + 8)

More information

Riemann is the Mann! (But Lebesgue may besgue to differ.)

Riemann is the Mann! (But Lebesgue may besgue to differ.) Riemnn is the Mnn! (But Lebesgue my besgue to differ.) Leo Livshits My 2, 2008 1 For finite intervls in R We hve seen in clss tht every continuous function f : [, b] R hs the property tht for every ɛ >

More information

INFORMATIONAL TEMPORARY EQUILIBRIA. J. S. Jordan * Discussion Paper No , April 1977

INFORMATIONAL TEMPORARY EQUILIBRIA. J. S. Jordan * Discussion Paper No , April 1977 INFORMATIONAL TEMPORARY EQUILIBRIA by J. S. Jordn * Dscusson Pper No. 77-89, Aprl 977 } * I m gretly ndebted to Professor S. Reter for severl stmultng dscussons, nd the orgnl descrpton (n [5J of the process

More information

Many-Body Calculations of the Isotope Shift

Many-Body Calculations of the Isotope Shift Mny-Body Clcultons of the Isotope Shft W. R. Johnson Mrch 11, 1 1 Introducton Atomc energy levels re commonly evluted ssumng tht the nucler mss s nfnte. In ths report, we consder correctons to tomc levels

More information

W. We shall do so one by one, starting with I 1, and we shall do it greedily, trying

W. We shall do so one by one, starting with I 1, and we shall do it greedily, trying Vitli covers 1 Definition. A Vitli cover of set E R is set V of closed intervls with positive length so tht, for every δ > 0 nd every x E, there is some I V with λ(i ) < δ nd x I. 2 Lemm (Vitli covering)

More information

Economics 101. Lecture 4 - Equilibrium and Efficiency

Economics 101. Lecture 4 - Equilibrium and Efficiency Economcs 0 Lecture 4 - Equlbrum and Effcency Intro As dscussed n the prevous lecture, we wll now move from an envronment where we looed at consumers mang decsons n solaton to analyzng economes full of

More information

Chapter 5 Supplemental Text Material R S T. ij i j ij ijk

Chapter 5 Supplemental Text Material R S T. ij i j ij ijk Chpter 5 Supplementl Text Mterl 5-. Expected Men Squres n the Two-fctor Fctorl Consder the two-fctor fxed effects model y = µ + τ + β + ( τβ) + ε k R S T =,,, =,,, k =,,, n gven s Equton (5-) n the textook.

More information

Review of Riemann Integral

Review of Riemann Integral 1 Review of Riemnn Integrl In this chpter we review the definition of Riemnn integrl of bounded function f : [, b] R, nd point out its limittions so s to be convinced of the necessity of more generl integrl.

More information

INTRODUCTION TO COMPLEX NUMBERS

INTRODUCTION TO COMPLEX NUMBERS INTRODUCTION TO COMPLEX NUMBERS The numers -4, -3, -, -1, 0, 1,, 3, 4 represent the negtve nd postve rel numers termed ntegers. As one frst lerns n mddle school they cn e thought of s unt dstnce spced

More information

NOTE AN INEQUALITY FOR KRUSKAL-MACAULAY FUNCTIONS

NOTE AN INEQUALITY FOR KRUSKAL-MACAULAY FUNCTIONS NOTE AN INEQUALITY FOR KRUSKAL-MACAULAY FUNCTIONS BERNARDO M. ÁBREGO, SILVIA FERNÁNDEZ-MERCHANT, AND BERNARDO LLANO Abstrct. Gven ntegers nd n, there s unque wy of wrtng n s n = n n... n so tht n <

More information

Econ107 Applied Econometrics Topic 3: Classical Model (Studenmund, Chapter 4)

Econ107 Applied Econometrics Topic 3: Classical Model (Studenmund, Chapter 4) I. Classcal Assumptons Econ7 Appled Econometrcs Topc 3: Classcal Model (Studenmund, Chapter 4) We have defned OLS and studed some algebrac propertes of OLS. In ths topc we wll study statstcal propertes

More information

Maximizing the number of nonnegative subsets

Maximizing the number of nonnegative subsets Maxmzng the number of nonnegatve subsets Noga Alon Hao Huang December 1, 213 Abstract Gven a set of n real numbers, f the sum of elements of every subset of sze larger than k s negatve, what s the maxmum

More information

Jens Siebel (University of Applied Sciences Kaiserslautern) An Interactive Introduction to Complex Numbers

Jens Siebel (University of Applied Sciences Kaiserslautern) An Interactive Introduction to Complex Numbers Jens Sebel (Unversty of Appled Scences Kserslutern) An Interctve Introducton to Complex Numbers 1. Introducton We know tht some polynoml equtons do not hve ny solutons on R/. Exmple 1.1: Solve x + 1= for

More information

Pyramid Algorithms for Barycentric Rational Interpolation

Pyramid Algorithms for Barycentric Rational Interpolation Pyrmd Algorthms for Brycentrc Rtonl Interpolton K Hormnn Scott Schefer Astrct We present new perspectve on the Floter Hormnn nterpolnt. Ths nterpolnt s rtonl of degree (n, d), reproduces polynomls of degree

More information

ψ ij has the eigenvalue

ψ ij has the eigenvalue Moller Plesset Perturbton Theory In Moller-Plesset (MP) perturbton theory one tes the unperturbed Hmltonn for n tom or molecule s the sum of the one prtcle Foc opertors H F() where the egenfunctons of

More information

Wars of attrition and all-pay auctions with stochastic competition

Wars of attrition and all-pay auctions with stochastic competition MPRA Munch Personl RePEc Archve Wrs of ttrton nd ll-py uctons wth stochstc competton Olver Bos Unversty Pnthéon-Asss, LEM 17. November 11 Onlne t http://mpr.ub.un-muenchen.de/3481/ MPRA Pper No. 3481,

More information

Exam 2, Mathematics 4701, Section ETY6 6:05 pm 7:40 pm, March 31, 2016, IH-1105 Instructor: Attila Máté 1

Exam 2, Mathematics 4701, Section ETY6 6:05 pm 7:40 pm, March 31, 2016, IH-1105 Instructor: Attila Máté 1 Exm, Mthemtics 471, Section ETY6 6:5 pm 7:4 pm, Mrch 1, 16, IH-115 Instructor: Attil Máté 1 17 copies 1. ) Stte the usul sufficient condition for the fixed-point itertion to converge when solving the eqution

More information

Polynomial Approximations for the Natural Logarithm and Arctangent Functions. Math 230

Polynomial Approximations for the Natural Logarithm and Arctangent Functions. Math 230 Polynomil Approimtions for the Nturl Logrithm nd Arctngent Functions Mth 23 You recll from first semester clculus how one cn use the derivtive to find n eqution for the tngent line to function t given

More information

arxiv:math/ v2 [math.ho] 16 Dec 2003

arxiv:math/ v2 [math.ho] 16 Dec 2003 rxiv:mth/0312293v2 [mth.ho] 16 Dec 2003 Clssicl Lebesgue Integrtion Theorems for the Riemnn Integrl Josh Isrlowitz 244 Ridge Rd. Rutherford, NJ 07070 jbi2@njit.edu Februry 1, 2008 Abstrct In this pper,

More information

6. Stochastic processes (2)

6. Stochastic processes (2) Contents Markov processes Brth-death processes Lect6.ppt S-38.45 - Introducton to Teletraffc Theory Sprng 5 Markov process Consder a contnuous-tme and dscrete-state stochastc process X(t) wth state space

More information

p (i.e., the set of all nonnegative real numbers). Similarly, Z will denote the set of all

p (i.e., the set of all nonnegative real numbers). Similarly, Z will denote the set of all th Prelmnry E 689 Lecture Notes by B. Yo 0. Prelmnry Notton themtcl Prelmnres It s ssumed tht the reder s fmlr wth the noton of set nd ts elementry oertons, nd wth some bsc logc oertors, e.g. x A : x s

More information

Sequences of Intuitionistic Fuzzy Soft G-Modules

Sequences of Intuitionistic Fuzzy Soft G-Modules Interntonl Mthemtcl Forum, Vol 13, 2018, no 12, 537-546 HIKARI Ltd, wwwm-hkrcom https://doorg/1012988/mf201881058 Sequences of Intutonstc Fuzzy Soft G-Modules Velyev Kemle nd Huseynov Afq Bku Stte Unversty,

More information

6. Stochastic processes (2)

6. Stochastic processes (2) 6. Stochastc processes () Lect6.ppt S-38.45 - Introducton to Teletraffc Theory Sprng 5 6. Stochastc processes () Contents Markov processes Brth-death processes 6. Stochastc processes () Markov process

More information

Riemann Sums and Riemann Integrals

Riemann Sums and Riemann Integrals Riemnn Sums nd Riemnn Integrls Jmes K. Peterson Deprtment of Biologicl Sciences nd Deprtment of Mthemticl Sciences Clemson University August 26, 2013 Outline 1 Riemnn Sums 2 Riemnn Integrls 3 Properties

More information

20. Mon, Oct. 13 What we have done so far corresponds roughly to Chapters 2 & 3 of Lee. Now we turn to Chapter 4. The first idea is connectedness.

20. Mon, Oct. 13 What we have done so far corresponds roughly to Chapters 2 & 3 of Lee. Now we turn to Chapter 4. The first idea is connectedness. 20. Mon, Oct. 13 What we have done so far corresponds roughly to Chapters 2 & 3 of Lee. Now we turn to Chapter 4. The frst dea s connectedness. Essentally, we want to say that a space cannot be decomposed

More information

Riemann Sums and Riemann Integrals

Riemann Sums and Riemann Integrals Riemnn Sums nd Riemnn Integrls Jmes K. Peterson Deprtment of Biologicl Sciences nd Deprtment of Mthemticl Sciences Clemson University August 26, 203 Outline Riemnn Sums Riemnn Integrls Properties Abstrct

More information

A Tri-Valued Belief Network Model for Information Retrieval

A Tri-Valued Belief Network Model for Information Retrieval December 200 A Tr-Vlued Belef Networ Model for Informton Retrevl Fernndo Ds-Neves Computer Scence Dept. Vrgn Polytechnc Insttute nd Stte Unversty Blcsburg, VA 24060. IR models t Combnng Evdence Grphcl

More information

Physics 121 Sample Common Exam 2 Rev2 NOTE: ANSWERS ARE ON PAGE 7. Instructions:

Physics 121 Sample Common Exam 2 Rev2 NOTE: ANSWERS ARE ON PAGE 7. Instructions: Physcs 121 Smple Common Exm 2 Rev2 NOTE: ANSWERS ARE ON PAGE 7 Nme (Prnt): 4 Dgt ID: Secton: Instructons: Answer ll 27 multple choce questons. You my need to do some clculton. Answer ech queston on the

More information

Electrochemical Thermodynamics. Interfaces and Energy Conversion

Electrochemical Thermodynamics. Interfaces and Energy Conversion CHE465/865, 2006-3, Lecture 6, 18 th Sep., 2006 Electrochemcl Thermodynmcs Interfces nd Energy Converson Where does the energy contrbuton F zϕ dn come from? Frst lw of thermodynmcs (conservton of energy):

More information

Inner Product. Euclidean Space. Orthonormal Basis. Orthogonal

Inner Product. Euclidean Space. Orthonormal Basis. Orthogonal Inner Product Defnton 1 () A Eucldean space s a fnte-dmensonal vector space over the reals R, wth an nner product,. Defnton 2 (Inner Product) An nner product, on a real vector space X s a symmetrc, blnear,

More information

Genericity of Critical Types

Genericity of Critical Types Genercty of Crtcal Types Y-Chun Chen Alfredo D Tllo Eduardo Fangold Syang Xong September 2008 Abstract Ely and Pesk 2008 offers an nsghtful characterzaton of crtcal types: a type s crtcal f and only f

More information

Chapter 2 Introduction to Algebra. Dr. Chih-Peng Li ( 李 )

Chapter 2 Introduction to Algebra. Dr. Chih-Peng Li ( 李 ) Chpter Introducton to Algebr Dr. Chh-Peng L 李 Outlne Groups Felds Bnry Feld Arthetc Constructon of Glos Feld Bsc Propertes of Glos Feld Coputtons Usng Glos Feld Arthetc Vector Spces Groups 3 Let G be set

More information

Module 3 LOSSY IMAGE COMPRESSION SYSTEMS. Version 2 ECE IIT, Kharagpur

Module 3 LOSSY IMAGE COMPRESSION SYSTEMS. Version 2 ECE IIT, Kharagpur Module 3 LOSSY IMAGE COMPRESSION SYSTEMS Verson ECE IIT, Kharagpur Lesson 6 Theory of Quantzaton Verson ECE IIT, Kharagpur Instructonal Objectves At the end of ths lesson, the students should be able to:

More information

Games of Threats. Elon Kohlberg Abraham Neyman. Working Paper

Games of Threats. Elon Kohlberg Abraham Neyman. Working Paper Games of Threats Elon Kohlberg Abraham Neyman Workng Paper 18-023 Games of Threats Elon Kohlberg Harvard Busness School Abraham Neyman The Hebrew Unversty of Jerusalem Workng Paper 18-023 Copyrght 2017

More information

Subjective Uncertainty Over Behavior Strategies: A Correction

Subjective Uncertainty Over Behavior Strategies: A Correction Subjectve Uncertanty Over Behavor Strateges: A Correcton The Harvard communty has made ths artcle openly avalable. Please share how ths access benefts you. Your story matters. Ctaton Publshed Verson Accessed

More information

p-adic Egyptian Fractions

p-adic Egyptian Fractions p-adic Egyptin Frctions Contents 1 Introduction 1 2 Trditionl Egyptin Frctions nd Greedy Algorithm 2 3 Set-up 3 4 p-greedy Algorithm 5 5 p-egyptin Trditionl 10 6 Conclusion 1 Introduction An Egyptin frction

More information

3.1 Expectation of Functions of Several Random Variables. )' be a k-dimensional discrete or continuous random vector, with joint PMF p (, E X E X1 E X

3.1 Expectation of Functions of Several Random Variables. )' be a k-dimensional discrete or continuous random vector, with joint PMF p (, E X E X1 E X Statstcs 1: Probablty Theory II 37 3 EPECTATION OF SEVERAL RANDOM VARIABLES As n Probablty Theory I, the nterest n most stuatons les not on the actual dstrbuton of a random vector, but rather on a number

More information

Lecture 1. Functional series. Pointwise and uniform convergence.

Lecture 1. Functional series. Pointwise and uniform convergence. 1 Introduction. Lecture 1. Functionl series. Pointwise nd uniform convergence. In this course we study mongst other things Fourier series. The Fourier series for periodic function f(x) with period 2π is

More information

A new construction of 3-separable matrices via an improved decoding of Macula s construction

A new construction of 3-separable matrices via an improved decoding of Macula s construction Dscrete Optmzaton 5 008 700 704 Contents lsts avalable at ScenceDrect Dscrete Optmzaton journal homepage: wwwelsevercom/locate/dsopt A new constructon of 3-separable matrces va an mproved decodng of Macula

More information

1 Online Learning and Regret Minimization

1 Online Learning and Regret Minimization 2.997 Decision-Mking in Lrge-Scle Systems My 10 MIT, Spring 2004 Hndout #29 Lecture Note 24 1 Online Lerning nd Regret Minimiztion In this lecture, we consider the problem of sequentil decision mking in

More information

APPENDIX A Some Linear Algebra

APPENDIX A Some Linear Algebra APPENDIX A Some Lnear Algebra The collecton of m, n matrces A.1 Matrces a 1,1,..., a 1,n A = a m,1,..., a m,n wth real elements a,j s denoted by R m,n. If n = 1 then A s called a column vector. Smlarly,

More information

Math 61CM - Solutions to homework 9

Math 61CM - Solutions to homework 9 Mth 61CM - Solutions to homework 9 Cédric De Groote November 30 th, 2018 Problem 1: Recll tht the left limit of function f t point c is defined s follows: lim f(x) = l x c if for ny > 0 there exists δ

More information

a b a In case b 0, a being divisible by b is the same as to say that

a b a In case b 0, a being divisible by b is the same as to say that Secton 6.2 Dvsblty among the ntegers An nteger a ε s dvsble by b ε f there s an nteger c ε such that a = bc. Note that s dvsble by any nteger b, snce = b. On the other hand, a s dvsble by only f a = :

More information

Lecture 3 ( ) (translated and slightly adapted from lecture notes by Martin Klazar)

Lecture 3 ( ) (translated and slightly adapted from lecture notes by Martin Klazar) Lecture 3 (5.3.2018) (trnslted nd slightly dpted from lecture notes by Mrtin Klzr) Riemnn integrl Now we define precisely the concept of the re, in prticulr, the re of figure U(, b, f) under the grph of

More information

SUMMER KNOWHOW STUDY AND LEARNING CENTRE

SUMMER KNOWHOW STUDY AND LEARNING CENTRE SUMMER KNOWHOW STUDY AND LEARNING CENTRE Indices & Logrithms 2 Contents Indices.2 Frctionl Indices.4 Logrithms 6 Exponentil equtions. Simplifying Surds 13 Opertions on Surds..16 Scientific Nottion..18

More information

Chapter Newton-Raphson Method of Solving a Nonlinear Equation

Chapter Newton-Raphson Method of Solving a Nonlinear Equation Chpter 0.04 Newton-Rphson Method o Solvng Nonlner Equton Ater redng ths chpter, you should be ble to:. derve the Newton-Rphson method ormul,. develop the lgorthm o the Newton-Rphson method,. use the Newton-Rphson

More information

Definition of Tracking

Definition of Tracking Trckng Defnton of Trckng Trckng: Generte some conclusons bout the moton of the scene, objects, or the cmer, gven sequence of mges. Knowng ths moton, predct where thngs re gong to project n the net mge,

More information

CHI-SQUARE DIVERGENCE AND MINIMIZATION PROBLEM

CHI-SQUARE DIVERGENCE AND MINIMIZATION PROBLEM CHI-SQUARE DIVERGENCE AND MINIMIZATION PROBLEM PRANESH KUMAR AND INDER JEET TANEJA Abstrct The mnmum dcrmnton nformton prncple for the Kullbck-Lebler cross-entropy well known n the lterture In th pper

More information

Lecture notes. Fundamental inequalities: techniques and applications

Lecture notes. Fundamental inequalities: techniques and applications Lecture notes Fundmentl nequltes: technques nd pplctons Mnh Hong Duong Mthemtcs Insttute, Unversty of Wrwck Eml: m.h.duong@wrwck.c.uk Jnury 4, 07 Abstrct Inequltes re ubqutous n Mthemtcs (nd n rel lfe.

More information

Advanced Calculus: MATH 410 Uniform Convergence of Functions Professor David Levermore 11 December 2015

Advanced Calculus: MATH 410 Uniform Convergence of Functions Professor David Levermore 11 December 2015 Advnced Clculus: MATH 410 Uniform Convergence of Functions Professor Dvid Levermore 11 December 2015 12. Sequences of Functions We now explore two notions of wht it mens for sequence of functions {f n

More information

Solution for Assignment 1 : Intro to Probability and Statistics, PAC learning

Solution for Assignment 1 : Intro to Probability and Statistics, PAC learning Solution for Assignment 1 : Intro to Probbility nd Sttistics, PAC lerning 10-701/15-781: Mchine Lerning (Fll 004) Due: Sept. 30th 004, Thursdy, Strt of clss Question 1. Bsic Probbility ( 18 pts) 1.1 (

More information

Advanced Calculus: MATH 410 Notes on Integrals and Integrability Professor David Levermore 17 October 2004

Advanced Calculus: MATH 410 Notes on Integrals and Integrability Professor David Levermore 17 October 2004 Advnced Clculus: MATH 410 Notes on Integrls nd Integrbility Professor Dvid Levermore 17 October 2004 1. Definite Integrls In this section we revisit the definite integrl tht you were introduced to when

More information

6 Roots of Equations: Open Methods

6 Roots of Equations: Open Methods HK Km Slghtly modfed 3//9, /8/6 Frstly wrtten t Mrch 5 6 Roots of Equtons: Open Methods Smple Fed-Pont Iterton Newton-Rphson Secnt Methods MATLAB Functon: fzero Polynomls Cse Study: Ppe Frcton Brcketng

More information

523 P a g e. is measured through p. should be slower for lesser values of p and faster for greater values of p. If we set p*

523 P a g e. is measured through p. should be slower for lesser values of p and faster for greater values of p. If we set p* R. Smpth Kumr, R. Kruthk, R. Rdhkrshnn / Interntonl Journl of Engneerng Reserch nd Applctons (IJERA) ISSN: 48-96 www.jer.com Vol., Issue 4, July-August 0, pp.5-58 Constructon Of Mxed Smplng Plns Indexed

More information

Chapter 1: Fundamentals

Chapter 1: Fundamentals Chpter 1: Fundmentls 1.1 Rel Numbers Types of Rel Numbers: Nturl Numbers: {1, 2, 3,...}; These re the counting numbers. Integers: {... 3, 2, 1, 0, 1, 2, 3,...}; These re ll the nturl numbers, their negtives,

More information

Math 8 Winter 2015 Applications of Integration

Math 8 Winter 2015 Applications of Integration Mth 8 Winter 205 Applictions of Integrtion Here re few importnt pplictions of integrtion. The pplictions you my see on n exm in this course include only the Net Chnge Theorem (which is relly just the Fundmentl

More information

COMPLEX NUMBER & QUADRATIC EQUATION

COMPLEX NUMBER & QUADRATIC EQUATION MCQ COMPLEX NUMBER & QUADRATIC EQUATION Syllus : Comple numers s ordered prs of rels, Representton of comple numers n the form + nd ther representton n plne, Argnd dgrm, lger of comple numers, modulus

More information

( dg. ) 2 dt. + dt. dt j + dh. + dt. r(t) dt. Comparing this equation with the one listed above for the length of see that

( dg. ) 2 dt. + dt. dt j + dh. + dt. r(t) dt. Comparing this equation with the one listed above for the length of see that Arc Length of Curves in Three Dimensionl Spce If the vector function r(t) f(t) i + g(t) j + h(t) k trces out the curve C s t vries, we cn mesure distnces long C using formul nerly identicl to one tht we

More information

An Introduction to Support Vector Machines

An Introduction to Support Vector Machines An Introducton to Support Vector Mchnes Wht s good Decson Boundry? Consder two-clss, lnerly seprble clssfcton problem Clss How to fnd the lne (or hyperplne n n-dmensons, n>)? Any de? Clss Per Lug Mrtell

More information

Available online through

Available online through Avlble ole through wwwmfo FIXED POINTS FOR NON-SELF MAPPINGS ON CONEX ECTOR METRIC SPACES Susht Kumr Moht* Deprtmet of Mthemtcs West Begl Stte Uverst Brst 4 PrgsNorth) Kolt 76 West Begl Id E-ml: smwbes@yhoo

More information

Chapter 5 : Continuous Random Variables

Chapter 5 : Continuous Random Variables STAT/MATH 395 A - PROBABILITY II UW Winter Qurter 216 Néhémy Lim Chpter 5 : Continuous Rndom Vribles Nottions. N {, 1, 2,...}, set of nturl numbers (i.e. ll nonnegtive integers); N {1, 2,...}, set of ll

More information

International Journal of Pure and Applied Sciences and Technology

International Journal of Pure and Applied Sciences and Technology Int. J. Pure Appl. Sc. Technol., () (), pp. 44-49 Interntonl Journl of Pure nd Appled Scences nd Technolog ISSN 9-67 Avlle onlne t www.jopst.n Reserch Pper Numercl Soluton for Non-Lner Fredholm Integrl

More information

FUNDAMENTALS OF REAL ANALYSIS by. III.1. Measurable functions. f 1 (

FUNDAMENTALS OF REAL ANALYSIS by. III.1. Measurable functions. f 1 ( FUNDAMNTALS OF RAL ANALYSIS by Doğn Çömez III. MASURABL FUNCTIONS AND LBSGU INTGRAL III.. Mesurble functions Hving the Lebesgue mesure define, in this chpter, we will identify the collection of functions

More information

Lectures - Week 4 Matrix norms, Conditioning, Vector Spaces, Linear Independence, Spanning sets and Basis, Null space and Range of a Matrix

Lectures - Week 4 Matrix norms, Conditioning, Vector Spaces, Linear Independence, Spanning sets and Basis, Null space and Range of a Matrix Lectures - Week 4 Matrx norms, Condtonng, Vector Spaces, Lnear Independence, Spannng sets and Bass, Null space and Range of a Matrx Matrx Norms Now we turn to assocatng a number to each matrx. We could

More information