Pricing for Coordination in Open Loop Differential Games

Size: px
Start display at page:

Download "Pricing for Coordination in Open Loop Differential Games"

Transcription

1 Preprnts of the 19th World Congress The Internatonal Federaton of Automatc Control Prcng for Coordnaton n Open Loop Dfferental Games Danel Calderone Lllan J. Ratlff S. Shankar Sastry Department of Electrcal Engneerng and Computer Scences, Unversty of Calforna, Berkeley, Berkeley, CA USA {danjc, ratlffl, sastry} at eecs.berkeley.edu Abstract: We provde a method for desgnng a quadratc prcng scheme to nduce a desred local Nash equlbrum n an open-loop dfferental game wth nonlnear dynamcs. In addton, we present condtons when the nduced equlbrum s a global equlbrum. The results are appled to the problem of nducng network managers to nvest n securty n a mult-network wth an epdemc model of the spread of malware. Keywords: Game Theory, prcng desgn, nonlnear systems 1. INTRODUCTION When resources are scarce, competton develops between self nterested agents. Game theory s an establshed technque for modelng ths nteracton, and t has emerged as an engneerng tool for analyss and synthess of systems comprsed of dynamcally coupled decson makng agents possessng competng nterests (Coogan et al. (2013),L and Marden (2011), Ratlff et al. (2012)). In such scenaros, the strateges chosen by the selfsh agents result n a soluton that s often neffcent from a socetal pont of vew. Ths motvates the desgn of coordnatng mechansms that nduce agents to play a Nash equlbrum wth desrable propertes, namely an equlbrum that s socally optmal. Engneerng problems n whch there are decson makng agents, ether wth competng nterests or dfferent nformaton sets, as well as a socal planner who s tasked wth coordnatng the agents are appearng more frequently n the lterature as technology s ntegrated nto nfrastructure (Oldewurtel et al. (2010)). It s mportant to accurately model these systems and develop control strateges accountng for the nterests of all the partcpatng agents whle meetng the organzatonal objectve whch may represent socal welfare or the common good. The nteracton of selfsh agents n mult-agent system may be cast as a dfferental game by modelng dynamcally coupled agents as strategc players. The coordnaton problem may then be cast as an optmzaton problem whereby a socal planner determnes a coordnaton mechansm ensurng the agents play the desred equlbrum. The coordnaton mechansms modfy the agents nomnal utlty functons thereby allowng for a socal planner to shape the strateges of the agents n order to meet a desred global objectve. In Ratlff et al. (2012); Calderone et al. (2013); Coogan et al. (2013) the problem of fndng prces to nduce a socally optmal Nash equlbrum n a lnear quadratc game s solved. In ths paper, we show that the result can be generalzed to open loop dfferental games wth non lnear dynamcs and possbly non convex costs that are separable n the state and the control. In partcular, we formulate the problem of desgnng prces to nduce a desred (socally optmal) equlbrum for games wth non lnear dynamcs and non convex costs as a feasblty problem. We show that f ths feasblty problem has a soluton, then the desred equlbrum s a local Nash equlbrum of the game resultng from mposton of prces on the players. Further, f the desred equlbrum s the unque equlbrum of the prcng nduced game, the desgned prces cause the agents to play the socally optmal soluton. If n addton the dynamcs are convex, then we provde an extended feasblty problem to desgn prces that force the socally optmal soluton to be the unque Nash equlbrum of the prcng nduced game. Fnally, we apply the theory to the problem of securty n mult networks whch s a rsng problem n the study of cyber physcal systems (Cardenas et al., 2008; Bloem et al., 2009; Alpcan and Başar, 2010). In ths example, we add an objectve functon and addtonal constrants to the prcng desgn feasblty problem to ensure a budget balanced soluton. The rest of the paper s organzed as follows. In Secton 2, we formulate the game. In Secton 3, we defne the prcng optmzaton problem and state our man results. n Secton 4, we consder the applcaton of desgnng prces to nduce nvestment n securty n a mult network system wth an epdemc model for the spread of malware. The prcng scheme guarantees that the network dynamcs remans stable meanng the spread of malware does not destablze the networks. In Secton 5, we summarze the contrbutons and dscuss future drectons. 2. AGENT GAME Consder a dynamc game wth n agents where the system dynamcs are gven by the general nonlnear ordnary dfferental equaton ẋ = f(t, x, u), x 0 (t) = x 0 (1) where Copyrght 2014 IFAC 9001

2 x(t) = [x 1 (t) T... x n (t) T ] T (2) and u(t) = [u 1 (t) T... u n (t) T ] T. (3) Each x (t) R n where n s the dmenson. The th agent has control over control nput u (t) R m where m s the dmenson of nput u (t) and has nomnal cost gven by J (x(t 0 ), u) = 1 q (t, x) + r (t, u) dt + q (t f, x) (4) 2 t 0 We suppress the dependence of the cost J on the ntal condton x(t 0 ) when t s clear from context. By an abuse of notaton, we let u denote the strategy over the horzon [t 0, t f ] and we drop the dependence on t n the control acton u (t) where t s clear from context. Each player s nterested n mnmzng ther cost J (u, u ) wth respect to ther choce varable u and where = {1,..., 1, + 1,..., n}. We restrct each agent s choce u to be an open loop control and we denote the space of open loop control strateges for agent by Γ. Gven the cost J (u, u ), the Hamltonan for the th player s gven by H (t, x, p, u, u ) = q (t, x) + r (t, u) + p (t) T f(t, x, u) (5) and the optmzed Hamltonan for the th player s gven by H (t, x, p ) = mn H (t, x, p, u, u ). (6) u Γ Note that the co state for each player s a vector p (t) R n for each t [t 0, t f ]. Let {u }n =1 be an open loop Nash equlbrum (ether local or global). Then, n equlbrum, the optmalty condtons for agent s optmzaton problem are ẋ(t) = H p (t, x, p, u ) (7) ṗ (t) T = H x (t, x, p, u ) (8) 0 = H (t, x, p, u ) (9) for all t [t 0, t f ] where and x(t 0 ) = x 0 and p (t f ) = q x (x (t f )). (10) H p (t, x, p, u ) = f(t, x, u ), (11) H x (t, x, p, u ) = q x (t, x ) + p T and H f x (t, x, u ), (12) (t, x, p, u ) = r (t, u ) + p T f (t, x, u ). (13) Equaton (7) s the state equaton, Equaton (8) s the co state equaton, and Equaton (9) s the nput statonarty condton. Defnton 1. A Nash equlbrum n the open loop dfferental game defned by the costs (4) and dynamcs (1) s a set of strateges {u [t 0, t f ]} n =1 such that for each {1,..., n} J (u, u ) J (u, u ) u Γ. (14) The above defnton can be nterpreted as sayng a set of strateges {u }n =1 s a Nash equlbrum f no player can decrease ther cost by unlaterally devatng from ther strategy n {u [t 0, t f ]} n =1. The problem of fndng an open loop Nash equlbrum for the game defned by (1) and (4) s to fnd a set of control strateges {u [t 0, t f ]} n =1 such that the nequalty (14) s satsfed for each {1,..., n}. In general, fndng open loop Nash equlbra of dynamc games s a dffcult problem. Some recent work has explored the characterzaton and computaton of local equlbra n contnuous games ncludng open loop dfferental games Ratlff et al. (2013). In Secton 4 we wll use the technques ntroduced n Ratlff et al. (2013) to compute the Nash equlbrum under the prcng scheme n order to valdate that the prcng scheme results n the desred behavor modfcaton. 3. PRICING DESIGN The prcng desgn problem s defned to be the optmzaton problem solved by the socal planner n whch she desgns prcng mechansms to nduce agents to use the desred equlbrum {u (t) = K (t)} n =1. We wll restrct ourselves to modfyng each player s cost by addng a prcng mechansm, P (t, u), that s composed of a quadratc term and a lnear term n the control nputs. Namely, we defne P (t, u) = u T (t)r (t)u(t) + c u(t) dt (15) t 0 where R (t) = R (t) T 0. Thus each player s new cost wth prcng s gven by J (u) = t 0 q (t, x) + r (u, t) + u T R (t)u + c u dt + q (t f, x(t f )). (16) For convenence, we wll partton R (t) and c (t) as follows: R (t) := [ R 1 (t) R n (t) ] and c (t) := [ c 1 (t) c n (t) ]. (17) The Hamltonan for the th agent generated under the prcng scheme s gven by H (t, x, p, u) = q (t, x) + r (t, u) + u(t) T R (t)u(t) + c (t)u + p T f(t, x, u) (18) and the optmzed Hamltonan under prcng s gven by H (t, x, p ) = mn H (t, x, p, u, u ). (19) u Γ The optmalty condtons under prcng are ẋ(t) = H p (t, x, p, u ) = f(t, x, u ) (20) ṗ T = H x (t, x, p, u ) = q x (t, x ) p T H (t, x, p, u ) = (R (t)) T u + c (t) + p T f x (t, x, u ) (21) f (t, x, u ) + r (t, u ) = 0. (22) Note that snce the costs are separable n the states and controls and snce we do not allow the prces to depend 9002

3 on the state, both the state equaton and the co state equaton are ndependent of the prcng mechansm. Thus, gven a set of desred controls, we can solve for the correspondng state trajectory, x(t) and co state trajectory p (t) on the tme nterval [t 0, t f ]. Then, use the state and co state to choose prces that satsfy the nput statonarty condton. Defne the desred equlbrum u = K(t) = [K 1 (t) K n (t)] T. (23) Further, let x (t) and p (t) denote the state and co state at tme t under the desred equlbrum control K(t) respectvely. At the desred equlbrum, the nput statonarty condton s r (t, K(t)) + R u (t) T K(t) + c (t) + p (t) T f (t, x (t), K(t)) = 0 (24) Rearrangng (24), we defne α (t) as follows: α (t) = R(t) T K(t) + c (t) = p (t) T f (t, x (t), K(t)) r (t, K(t)) (25) for each {1,..., n}. Note that α (t) s completely known. Thus, we want to fnd R (t) and c (t) to satsfy (25). Thus desgnng prces to make K(t) a local Nash equlbrum amounts to solvng the feasblty problem defned below n Equaton (26). R(t)K(t) + c(t) = α(t) t (26) where R(t) = [R1(t) 1 Rn(t)] n T (27) c(t) = [c 1 1(t) c n n(t)] T (28) α(t) = [α 1 (t) α n (t)] T. (29) We wll use the notaton R[t 0, t f ] and c[t 0, t f ] to denote R(t) and c(t) for each t [t 0, t f ]. We can summarze the above results n the followng theorem. Theorem 1. Consder the game defned by nomnal agent costs (4) and dynamcs (1). Let {u [t 0, t f ]} n =1 be the desred Nash equlbrum. If there exsts a soluton (R[t 0, t f ], c[t 0, t f ]) (30) to the feasblty problem defned n Equaton (26), then the desred soluton {u [t 0, t f ]} n =1 s a local Nash equlbrum to the prcng nduced game defned by costs (16) and dynamcs (1). In the case that the desred soluton s the unque Nash equlbrum to the nduced game, we get the followng result. Corollary 1. Consder the game defned by nomnal agent costs (4) and dynamcs (1). Let {u [t 0, t f ]} n =1 be the desred Nash equlbrum. If there exsts a soluton (R[t 0, t f ], c[t 0, t f ]) (31) to the feasblty problem defned n Equaton (26) and the desred Nash equlbrum s unque n the resultng game defned by costs (16) and dynamcs (1), then the prces (R[t 0, t f ], c[t 0, t f ]) nduce the agents to play {u [t 0, t f ]} n = Dynamcs Convex n the State and Control In the case that the desred equlbrum s unque, the prcng mechansms are guaranteed to enforce the desred equlbrum strateges. Let us recall the noton of strct dagonal convexty ntroduced by Rosen n Rosen (1965) and then extended to the nfnte dmensonal case n Haure and Moresno (2001). Defnton 2. A functon L (x, u, t, p ) s dagonally strctly convex n (x, u) f for all ū, û, x, and ˆx we have (ū û ) T Å L (x, ū, t, p ) L (x, û, t, p ) ( x ˆx ) T Å L x ( x, u, t, p ) L x (ˆx, u, t, p ) ã ã > 0 (32) The followng lemma and theorem provde condtons under whch a Nash equlbrum of an open loop dfferental game s unque. Lemma 1. Assume that L (x, u) s convex n (x, u) and assume that the total runnng cost L(x, u) = L (x, u) (33) s dagonally strctly convex n (x, u). Further, assume that the dynamcs f(t, x, u) are convex n x and u. Then, the combned Hamltonan H(t, x, p) = H (t, x, p ) (34) s dagonally strctly convex n x and concave n p. Theorem 2. If the combned Hamltonan H(t, x, p) s dagonally strctly convex n x and concave n p, then the open loop Nash equlbrum s unque. Lemma 1 s a modfed verson of Lemma 2.1 n Haure and Moresno (2001). Theorem 2 s stated n Haure and Moresno (2001) and ts proof can be found n Carlson and Haure (1996) usng our verson of the lemma. If the costs under prcng satsfy the assumptons of Lemma 1, the desred equlbrum s the unque Nash equlbrum of the open loop dfferental game nduced through prcng. Defne the total runnng cost L (x, u, t, p ) (35) where L (x, u, t, p ) = q (x, t) + r (u, t) + u T R (t)u + c (t)u(t). (36) Assumpton 1. q (x, t) + r (u, t) s dagonally strctly convex n (x, u). Provded Assumpton 1, by Lemma 1 and Theorem 2, we only need to ensure that the prcng mechansm s dagonally strctly convex n u,.e. we need to enforce (û(t) ū(t)) T R(t)(û(t) ū(t)) > 0 (37) for all û and ū and for each t [t 0, t f ]. Thus n order to ensure dagonal strct convexty n u, we smply need to ensure that the symmetrc component of R(t),.e. R(t) + R(t) T, s postve defnte for all t. 9003

4 ß R(t)K(t) + c(t) = α(t) t R(t) + R(t) T 0 t (38) We have the followng result. Theorem 3. Consder the game defned by nomnal agent costs (4) and dynamcs (1). Suppose that f(t, x, u) s convex n x and u. Let {u [t 0, t f ]} n =1 be the desred Nash equlbrum. Then, f there exsts a soluton (R[t 0, t f ], c[t 0, t f ]) (39) to the feasblty problem defned n Equaton (38), then the prces (R[t 0, t f ], c[t 0, t f ]) nduce the agents to play {u [t 0, t f ]} n =1 to the game defned by (16) and dynamcs (1). Further, the desred Nash equlbrum s the unque equlbrum n the prcng nduced game. 4. PRICING IN MULTI NETWORKS We use the epdemc model for the spread of malware n a mult network ntroduced n Bloem et al. (2009). Selfspreadng attacks on computer networks are expensve owng to the damage they cause and the securty nvestment requred to defend aganst them. The socal planner s goal s to desgn prcng mechansms that coordnate the networks so that the overall mult-network s stablzed. Suppose that we have n networks wth N nodes n the th network and let x (t) R denote the number of nfected hosts n a network where hosts can be fractonally nfected. Let u (t) be the malware removal rate for network, α be the cross network parwse rate of nfecton, and β be the parwse rate of nfecton wthn networks. In general, computers wthn a network are more lkely to communcate wth one another than across networks; hence, we assume β > α. The spread of malware s then captured n the followng epdemc model: ẋ (t) = β(n x (t))x (t)+ α(n x (t))x j (t) u (t) j=1,j (40) Each network ndependently tres to choose u so that the th network s stablzed. For each network n the mult network, we consder a cost that s quadratc n the state,.e. the number of nfected hosts, and quadratc n the control,.e. the patchng rate. The nomnal cost for network s J (u) = 0 x T Q x + u T M u dt. (41) where Q and M are the cost of an nfected network host and the cost of the mplemented patchng response respectvely. The socal planner desgns prcng mechansms to coordnate the networks by nducng them to choose a desred control acton whch stablzes the entre multnetwork. We consder a group of sx networks wth N 1 = 3500, N 2 = 500, N 3 = 2000, N 4 = 1000, N 5 = 500, and N 6 = For each network, we take Q to be a dagonal matrx wth random postve entres where the (, )th element of Q s larger than the others. The M matrces are chosen to be 0 except for the (, )th element whch s 1. The rato between Q (, ) and M (, ) s 10 to 1. We scale the tme horzon to be over the nterval [0, 1] and take the ntal number of nfected nodes n each network to be half of the total number of nodes. As n Bloem et al. (2009), we take β = We set α = 2 3 β. Usng the dscretzaton scheme for optmal control problems descrbed n Chapter 4 of Polak (1997), we compute a centralzed soluton usng the sum of all the agents costs and standard nonlnear programmng technques. In general, ths only gves us a local optmum to the centralzed problem, but we wll see that t does mprove the performance of the system as compared to the Nash equlbrum. From the centralzed soluton, we determne a desred set of controls, K(t) = [K 1 (t) K n (t)] T. In order to desgn prces, the socal planners solves a modfed verson of the feasblty problem outlned n (38) at each tme step. Snce the nomnal costs are quadratc n the control, we replace each M wth R so that the new cost for each player becomes J (u) = 0 x T Q x + u T R (t)u + c (t)u dt. (42) In ths case, Equaton (25) becomes ᾱ (t) = R(t) T K(t) + c (t) = p (t) T f (t, x (t), K(t)) In addton, we add an objectve and several more constrants to make the problem budget balanced. Our fnal optmzaton problem s gven by mn {R (t),c (t)} n =1 R (t) M + c (t) (43) =1 subject to: R(t)K(t) + c(t) = ᾱ(t) (44) R(t) + R(t) T 0 (45) R (t) M 0 (46) =1 =1 R (t) = R (t) T 0, (47) at each tme step. Recall that R(t) = [R 1 1(t) R n n(t)] T. Equaton (46) forces the sum of the quadratc components of the prces to be greater than the sum of the nomnal quadratc components. Gven ths constrant, (43) seeks to make the costs wth prces as close as possble to the nomnal costs. Usng the same dscretzaton scheme as n the centralzed problem, we numercally approxmate local Nash equlbra under both the nomnal costs and the costs wth prcng usng the steepest descent algorthm presented n Ratlff et al. (2013). Snce the dynamcs are not convex n the state, we cannot guarantee global unqueness of the nduced equlbrum; however by checkng the 2nd-order suffcent condtons presented n Ratlff et al. (2013), we can show that the Nash under prcng s an solated local Nash. Thus control sgnals ntalzed close to the equlbrum wll converge to t under the steepest descent algorthm. By solvng the prcng problem, we make the centralzed soluton a local Nash equlbrum of the game wth prces. Moreover, we fnd that a wde varety of ntalzatons for the controls actually converge to the desred equlbrum. Computaton of basns of attracton of equlbra for ths problem and other general nonlnear problems s left as future work. 9004

5 It should be noted that the 2nd-order suffcent condtons for solated local Nash equlbra are only applcable to fnte dmensonal problems. Thus we can not guarantee that we nduce an solated equlbra of the actual nfnte dmensonal optmzaton problem but only of the fntedmensonal dscretzed problem. Fgure 1 shows the control nputs and Fgure 2 shows the state trajectores for the nomnal Nash equlbrum, the centralzed optmal soluton, and the Nash equlbrum under prcng. centralzed problem at each tme step. Each ndvdual player s runnng cost wth prces s not guaranteed to be equal to ther porton of the socal cost, however. As an example n Fgure 4, we plot the runnng costs of players 1 and 6. Fg. 3. Sum of ndvdual runnng costs at the nomnal Nash, socal optmum, and prcng nduced Nash. The fact that the socal optmum cost and prcng nduced Nash cost are equal means that we can acheve budget balance. Fg. 1. Control sgnals at the nomnal Nash equlbrum, the centralzed optmum, and prcng nduced Nash equlbrum. Fg. 2. Number of nfected nodes n each network over tme. Note that the nomnal Nash strateges do not elmnate all nfected nodes where as the socally optmal strateges and prcng nduced Nash strateges do. Fgure 3 compares the sum of the runnng costs. The centralzed soluton as well as the Nash under prcng reduces the total cost to the system by 8.2%. We also see that we are able to force the sum of all the runnng costs wth prces to be equal to the runnng cost of the Fg. 4. Indvdual runnng costs for players 1 and 6. Though the sum of all the runnng costs s the same as the socal optmum under prcng, each ndvdual player s runnng cost s dfferent from ther porton of the socal cost. 5. CONCLUSION In summary, we have formulated a feasblty problem for fndng quadratc and lnear prces to nduce a socally optmal local Nash equlbra n the context of open-loop dfferental games wth non-lnear dynamcs and general nomnal costs. In addton, we have shown that under specal condtons on the dynamcs and prces, the nduced 9005

6 equlbrum s the global equlbrum of the game wth prces. We apply these technques to the problem of ncentvzng network managers to nvest n securty to prevent the spread of epdemcs. In ths partcular example, we are able to desgn budget balanced prces that make the socally optmal controls an solated local Nash equlbra. We are currently nvestgatng computaton of basns of attracton for the equlbra of the nduced game. We are also studyng the stablty of the equlbra as well as condtons to ensure unqueness of equlbra n non-convex games. REFERENCES Alpcan, T. and Başar, T. (2010). Network securty: A decson and game-theoretc approach. Cambrdge Unversty Press. Bloem, M., Alpcan, T., and Başar, T. (2009). Optmal and robust epdemc response for multple networks. Control Engneerng Practce, 17(5), Calderone, D., Ratlff, L.J., and Sastry, S.S. (2013). Prcng desgn for robustness n lnear-quadratc dynamc games. In In the Proceedngs of the 52rd Annual IEEE Conference on Decson and Control. Cardenas, A., Amn, S., and Sastry, S. (2008). Secure control: Towards survvable cyber-physcal systems. In Dstrbuted Computng Systems Workshops, ICDCS th Internatonal Conference on, Carlson, D. and Haure, A. (1996). A turnpke theory for nfnte-horzon open-loop compettve processes. SIAM Journal on Control and Optmzaton, 34(4), Coogan, S., Ratlff, L., Calderone, D., Tomln, C., and Sastry, S.S. (2013). Energy management va prcng n LQ dynamc games. In Amercan Control Conference. Haure, A. and Moresno, F. (2001). Computaton of S- adapted Equlbra n Pecewse Determnstc Games va Stochastc Programmng Methods. Brkhäuser Boston. L, N. and Marden, J.R. (2011). Desgnng games for dstrbuted optmzaton. In Proceedngs of the 50th IEEE Conference on Decson and Control and European Control Conference, Oldewurtel, F., Ulbg, A., Parso, A., Andersson, G., and Morar, M. (2010). Reducng peak electrcty demand n buldng clmate control usng real-tme prcng and model predctve control. In Proceedngs of the 49th IEEE Conference on Decson and Control, Polak, E. (1997). Optmzaton: algorthms and consstent approxmatons. Sprnger New York. Ratlff, L.J., Burden, S.A., and Sastry, S.S. (2013). Characterzaton and computaton of local nash equlbra n contnuous games. In Proceedngs of the 51st Annual Allerton Conference on Communcaton, Control, and Computng. Ratlff, L., Coogan, S., Calderone, D., and Sastry, S. (2012). Prcng n lnear-quadratc dynamc games. In Proceedngs of the 50th Annual Allerton Conference on Communcaton, Control, and Computng, Rosen, J.B. (1965). Exstence and unqueness of equlbrum ponts for concave n-person games. Econometrca, 33(3),

The Second Anti-Mathima on Game Theory

The Second Anti-Mathima on Game Theory The Second Ant-Mathma on Game Theory Ath. Kehagas December 1 2006 1 Introducton In ths note we wll examne the noton of game equlbrum for three types of games 1. 2-player 2-acton zero-sum games 2. 2-player

More information

Numerical Heat and Mass Transfer

Numerical Heat and Mass Transfer Master degree n Mechancal Engneerng Numercal Heat and Mass Transfer 06-Fnte-Dfference Method (One-dmensonal, steady state heat conducton) Fausto Arpno f.arpno@uncas.t Introducton Why we use models and

More information

MMA and GCMMA two methods for nonlinear optimization

MMA and GCMMA two methods for nonlinear optimization MMA and GCMMA two methods for nonlnear optmzaton Krster Svanberg Optmzaton and Systems Theory, KTH, Stockholm, Sweden. krlle@math.kth.se Ths note descrbes the algorthms used n the author s 2007 mplementatons

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

Pricing in Linear-Quadratic Dynamic Games

Pricing in Linear-Quadratic Dynamic Games Prcng n Lnear-Quadratc Dynamc Games Lllan J. Ratlff, Samuel Coogan, Danel Calderone, and S. Shankar Sastry Abstract We nvestgate the use of prcng mechansms as a means to acheve a desred feedback control

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

princeton univ. F 17 cos 521: Advanced Algorithm Design Lecture 7: LP Duality Lecturer: Matt Weinberg

princeton univ. F 17 cos 521: Advanced Algorithm Design Lecture 7: LP Duality Lecturer: Matt Weinberg prnceton unv. F 17 cos 521: Advanced Algorthm Desgn Lecture 7: LP Dualty Lecturer: Matt Wenberg Scrbe: LP Dualty s an extremely useful tool for analyzng structural propertes of lnear programs. Whle there

More information

Lecture 12: Discrete Laplacian

Lecture 12: Discrete Laplacian Lecture 12: Dscrete Laplacan Scrbe: Tanye Lu Our goal s to come up wth a dscrete verson of Laplacan operator for trangulated surfaces, so that we can use t n practce to solve related problems We are mostly

More information

U.C. Berkeley CS294: Spectral Methods and Expanders Handout 8 Luca Trevisan February 17, 2016

U.C. Berkeley CS294: Spectral Methods and Expanders Handout 8 Luca Trevisan February 17, 2016 U.C. Berkeley CS94: Spectral Methods and Expanders Handout 8 Luca Trevsan February 7, 06 Lecture 8: Spectral Algorthms Wrap-up In whch we talk about even more generalzatons of Cheeger s nequaltes, and

More information

Chapter 5. Solution of System of Linear Equations. Module No. 6. Solution of Inconsistent and Ill Conditioned Systems

Chapter 5. Solution of System of Linear Equations. Module No. 6. Solution of Inconsistent and Ill Conditioned Systems Numercal Analyss by Dr. Anta Pal Assstant Professor Department of Mathematcs Natonal Insttute of Technology Durgapur Durgapur-713209 emal: anta.bue@gmal.com 1 . Chapter 5 Soluton of System of Lnear Equatons

More information

Lecture Notes on Linear Regression

Lecture Notes on Linear Regression Lecture Notes on Lnear Regresson Feng L fl@sdueducn Shandong Unversty, Chna Lnear Regresson Problem In regresson problem, we am at predct a contnuous target value gven an nput feature vector We assume

More information

Welfare Properties of General Equilibrium. What can be said about optimality properties of resource allocation implied by general equilibrium?

Welfare Properties of General Equilibrium. What can be said about optimality properties of resource allocation implied by general equilibrium? APPLIED WELFARE ECONOMICS AND POLICY ANALYSIS Welfare Propertes of General Equlbrum What can be sad about optmalty propertes of resource allocaton mpled by general equlbrum? Any crteron used to compare

More information

Problem Set 9 Solutions

Problem Set 9 Solutions Desgn and Analyss of Algorthms May 4, 2015 Massachusetts Insttute of Technology 6.046J/18.410J Profs. Erk Demane, Srn Devadas, and Nancy Lynch Problem Set 9 Solutons Problem Set 9 Solutons Ths problem

More information

LOW BIAS INTEGRATED PATH ESTIMATORS. James M. Calvin

LOW BIAS INTEGRATED PATH ESTIMATORS. James M. Calvin Proceedngs of the 007 Wnter Smulaton Conference S G Henderson, B Bller, M-H Hseh, J Shortle, J D Tew, and R R Barton, eds LOW BIAS INTEGRATED PATH ESTIMATORS James M Calvn Department of Computer Scence

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

COS 521: Advanced Algorithms Game Theory and Linear Programming

COS 521: Advanced Algorithms Game Theory and Linear Programming COS 521: Advanced Algorthms Game Theory and Lnear Programmng Moses Charkar February 27, 2013 In these notes, we ntroduce some basc concepts n game theory and lnear programmng (LP). We show a connecton

More information

CS294 Topics in Algorithmic Game Theory October 11, Lecture 7

CS294 Topics in Algorithmic Game Theory October 11, Lecture 7 CS294 Topcs n Algorthmc Game Theory October 11, 2011 Lecture 7 Lecturer: Chrstos Papadmtrou Scrbe: Wald Krchene, Vjay Kamble 1 Exchange economy We consder an exchange market wth m agents and n goods. Agent

More information

Feature Selection: Part 1

Feature Selection: Part 1 CSE 546: Machne Learnng Lecture 5 Feature Selecton: Part 1 Instructor: Sham Kakade 1 Regresson n the hgh dmensonal settng How do we learn when the number of features d s greater than the sample sze n?

More information

n α j x j = 0 j=1 has a nontrivial solution. Here A is the n k matrix whose jth column is the vector for all t j=0

n α j x j = 0 j=1 has a nontrivial solution. Here A is the n k matrix whose jth column is the vector for all t j=0 MODULE 2 Topcs: Lnear ndependence, bass and dmenson We have seen that f n a set of vectors one vector s a lnear combnaton of the remanng vectors n the set then the span of the set s unchanged f that vector

More information

Hongyi Miao, College of Science, Nanjing Forestry University, Nanjing ,China. (Received 20 June 2013, accepted 11 March 2014) I)ϕ (k)

Hongyi Miao, College of Science, Nanjing Forestry University, Nanjing ,China. (Received 20 June 2013, accepted 11 March 2014) I)ϕ (k) ISSN 1749-3889 (prnt), 1749-3897 (onlne) Internatonal Journal of Nonlnear Scence Vol.17(2014) No.2,pp.188-192 Modfed Block Jacob-Davdson Method for Solvng Large Sparse Egenproblems Hongy Mao, College of

More information

ECE559VV Project Report

ECE559VV Project Report ECE559VV Project Report (Supplementary Notes Loc Xuan Bu I. MAX SUM-RATE SCHEDULING: THE UPLINK CASE We have seen (n the presentaton that, for downlnk (broadcast channels, the strategy maxmzng the sum-rate

More information

Transfer Functions. Convenient representation of a linear, dynamic model. A transfer function (TF) relates one input and one output: ( ) system

Transfer Functions. Convenient representation of a linear, dynamic model. A transfer function (TF) relates one input and one output: ( ) system Transfer Functons Convenent representaton of a lnear, dynamc model. A transfer functon (TF) relates one nput and one output: x t X s y t system Y s The followng termnology s used: x y nput output forcng

More information

A SURVEY OF PROPERTIES OF FINITE HORIZON DIFFERENTIAL GAMES UNDER ISAACS CONDITION. Contents

A SURVEY OF PROPERTIES OF FINITE HORIZON DIFFERENTIAL GAMES UNDER ISAACS CONDITION. Contents A SURVEY OF PROPERTIES OF FINITE HORIZON DIFFERENTIAL GAMES UNDER ISAACS CONDITION BOTAO WU Abstract. In ths paper, we attempt to answer the followng questons about dfferental games: 1) when does a two-player,

More information

Lecture 21: Numerical methods for pricing American type derivatives

Lecture 21: Numerical methods for pricing American type derivatives Lecture 21: Numercal methods for prcng Amercan type dervatves Xaoguang Wang STAT 598W Aprl 10th, 2014 (STAT 598W) Lecture 21 1 / 26 Outlne 1 Fnte Dfference Method Explct Method Penalty Method (STAT 598W)

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

EEL 6266 Power System Operation and Control. Chapter 3 Economic Dispatch Using Dynamic Programming

EEL 6266 Power System Operation and Control. Chapter 3 Economic Dispatch Using Dynamic Programming EEL 6266 Power System Operaton and Control Chapter 3 Economc Dspatch Usng Dynamc Programmng Pecewse Lnear Cost Functons Common practce many utltes prefer to represent ther generator cost functons as sngle-

More information

Convexity preserving interpolation by splines of arbitrary degree

Convexity preserving interpolation by splines of arbitrary degree Computer Scence Journal of Moldova, vol.18, no.1(52), 2010 Convexty preservng nterpolaton by splnes of arbtrary degree Igor Verlan Abstract In the present paper an algorthm of C 2 nterpolaton of dscrete

More information

U.C. Berkeley CS294: Beyond Worst-Case Analysis Luca Trevisan September 5, 2017

U.C. Berkeley CS294: Beyond Worst-Case Analysis Luca Trevisan September 5, 2017 U.C. Berkeley CS94: Beyond Worst-Case Analyss Handout 4s Luca Trevsan September 5, 07 Summary of Lecture 4 In whch we ntroduce semdefnte programmng and apply t to Max Cut. Semdefnte Programmng Recall that

More information

Generalized Linear Methods

Generalized Linear Methods Generalzed Lnear Methods 1 Introducton In the Ensemble Methods the general dea s that usng a combnaton of several weak learner one could make a better learner. More formally, assume that we have a set

More information

Computing Correlated Equilibria in Multi-Player Games

Computing Correlated Equilibria in Multi-Player Games Computng Correlated Equlbra n Mult-Player Games Chrstos H. Papadmtrou Presented by Zhanxang Huang December 7th, 2005 1 The Author Dr. Chrstos H. Papadmtrou CS professor at UC Berkley (taught at Harvard,

More information

Outline and Reading. Dynamic Programming. Dynamic Programming revealed. Computing Fibonacci. The General Dynamic Programming Technique

Outline and Reading. Dynamic Programming. Dynamic Programming revealed. Computing Fibonacci. The General Dynamic Programming Technique Outlne and Readng Dynamc Programmng The General Technque ( 5.3.2) -1 Knapsac Problem ( 5.3.3) Matrx Chan-Product ( 5.3.1) Dynamc Programmng verson 1.4 1 Dynamc Programmng verson 1.4 2 Dynamc Programmng

More information

Composite Hypotheses testing

Composite Hypotheses testing Composte ypotheses testng In many hypothess testng problems there are many possble dstrbutons that can occur under each of the hypotheses. The output of the source s a set of parameters (ponts n a parameter

More information

How Strong Are Weak Patents? Joseph Farrell and Carl Shapiro. Supplementary Material Licensing Probabilistic Patents to Cournot Oligopolists *

How Strong Are Weak Patents? Joseph Farrell and Carl Shapiro. Supplementary Material Licensing Probabilistic Patents to Cournot Oligopolists * How Strong Are Weak Patents? Joseph Farrell and Carl Shapro Supplementary Materal Lcensng Probablstc Patents to Cournot Olgopolsts * September 007 We study here the specal case n whch downstream competton

More information

THE GUARANTEED COST CONTROL FOR UNCERTAIN LARGE SCALE INTERCONNECTED SYSTEMS

THE GUARANTEED COST CONTROL FOR UNCERTAIN LARGE SCALE INTERCONNECTED SYSTEMS Copyrght 22 IFAC 5th rennal World Congress, Barcelona, Span HE GUARANEED COS CONROL FOR UNCERAIN LARGE SCALE INERCONNECED SYSEMS Hroak Mukadan Yasuyuk akato Yoshyuk anaka Koch Mzukam Faculty of Informaton

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

The Minimum Universal Cost Flow in an Infeasible Flow Network

The Minimum Universal Cost Flow in an Infeasible Flow Network Journal of Scences, Islamc Republc of Iran 17(2): 175-180 (2006) Unversty of Tehran, ISSN 1016-1104 http://jscencesutacr The Mnmum Unversal Cost Flow n an Infeasble Flow Network H Saleh Fathabad * M Bagheran

More information

STAT 309: MATHEMATICAL COMPUTATIONS I FALL 2018 LECTURE 16

STAT 309: MATHEMATICAL COMPUTATIONS I FALL 2018 LECTURE 16 STAT 39: MATHEMATICAL COMPUTATIONS I FALL 218 LECTURE 16 1 why teratve methods f we have a lnear system Ax = b where A s very, very large but s ether sparse or structured (eg, banded, Toepltz, banded plus

More information

Dynamic Programming. Preview. Dynamic Programming. Dynamic Programming. Dynamic Programming (Example: Fibonacci Sequence)

Dynamic Programming. Preview. Dynamic Programming. Dynamic Programming. Dynamic Programming (Example: Fibonacci Sequence) /24/27 Prevew Fbonacc Sequence Longest Common Subsequence Dynamc programmng s a method for solvng complex problems by breakng them down nto smpler sub-problems. It s applcable to problems exhbtng the propertes

More information

Lecture 20: Lift and Project, SDP Duality. Today we will study the Lift and Project method. Then we will prove the SDP duality theorem.

Lecture 20: Lift and Project, SDP Duality. Today we will study the Lift and Project method. Then we will prove the SDP duality theorem. prnceton u. sp 02 cos 598B: algorthms and complexty Lecture 20: Lft and Project, SDP Dualty Lecturer: Sanjeev Arora Scrbe:Yury Makarychev Today we wll study the Lft and Project method. Then we wll prove

More information

Introduction. - The Second Lyapunov Method. - The First Lyapunov Method

Introduction. - The Second Lyapunov Method. - The First Lyapunov Method Stablty Analyss A. Khak Sedgh Control Systems Group Faculty of Electrcal and Computer Engneerng K. N. Toos Unversty of Technology February 2009 1 Introducton Stablty s the most promnent characterstc of

More information

Finite Element Modelling of truss/cable structures

Finite Element Modelling of truss/cable structures Pet Schreurs Endhoven Unversty of echnology Department of Mechancal Engneerng Materals echnology November 3, 214 Fnte Element Modellng of truss/cable structures 1 Fnte Element Analyss of prestressed structures

More information

A new Approach for Solving Linear Ordinary Differential Equations

A new Approach for Solving Linear Ordinary Differential Equations , ISSN 974-57X (Onlne), ISSN 974-5718 (Prnt), Vol. ; Issue No. 1; Year 14, Copyrght 13-14 by CESER PUBLICATIONS A new Approach for Solvng Lnear Ordnary Dfferental Equatons Fawz Abdelwahd Department of

More information

Notes on Frequency Estimation in Data Streams

Notes on Frequency Estimation in Data Streams Notes on Frequency Estmaton n Data Streams In (one of) the data streamng model(s), the data s a sequence of arrvals a 1, a 2,..., a m of the form a j = (, v) where s the dentty of the tem and belongs to

More information

Online Appendix. t=1 (p t w)q t. Then the first order condition shows that

Online Appendix. t=1 (p t w)q t. Then the first order condition shows that Artcle forthcomng to ; manuscrpt no (Please, provde the manuscrpt number!) 1 Onlne Appendx Appendx E: Proofs Proof of Proposton 1 Frst we derve the equlbrum when the manufacturer does not vertcally ntegrate

More information

Interactive Bi-Level Multi-Objective Integer. Non-linear Programming Problem

Interactive Bi-Level Multi-Objective Integer. Non-linear Programming Problem Appled Mathematcal Scences Vol 5 0 no 65 3 33 Interactve B-Level Mult-Objectve Integer Non-lnear Programmng Problem O E Emam Department of Informaton Systems aculty of Computer Scence and nformaton Helwan

More information

The Quadratic Trigonometric Bézier Curve with Single Shape Parameter

The Quadratic Trigonometric Bézier Curve with Single Shape Parameter J. Basc. Appl. Sc. Res., (3541-546, 01 01, TextRoad Publcaton ISSN 090-4304 Journal of Basc and Appled Scentfc Research www.textroad.com The Quadratc Trgonometrc Bézer Curve wth Sngle Shape Parameter Uzma

More information

Perron Vectors of an Irreducible Nonnegative Interval Matrix

Perron Vectors of an Irreducible Nonnegative Interval Matrix Perron Vectors of an Irreducble Nonnegatve Interval Matrx Jr Rohn August 4 2005 Abstract As s well known an rreducble nonnegatve matrx possesses a unquely determned Perron vector. As the man result of

More information

Yong Joon Ryang. 1. Introduction Consider the multicommodity transportation problem with convex quadratic cost function. 1 2 (x x0 ) T Q(x x 0 )

Yong Joon Ryang. 1. Introduction Consider the multicommodity transportation problem with convex quadratic cost function. 1 2 (x x0 ) T Q(x x 0 ) Kangweon-Kyungk Math. Jour. 4 1996), No. 1, pp. 7 16 AN ITERATIVE ROW-ACTION METHOD FOR MULTICOMMODITY TRANSPORTATION PROBLEMS Yong Joon Ryang Abstract. The optmzaton problems wth quadratc constrants often

More information

Additional Codes using Finite Difference Method. 1 HJB Equation for Consumption-Saving Problem Without Uncertainty

Additional Codes using Finite Difference Method. 1 HJB Equation for Consumption-Saving Problem Without Uncertainty Addtonal Codes usng Fnte Dfference Method Benamn Moll 1 HJB Equaton for Consumpton-Savng Problem Wthout Uncertanty Before consderng the case wth stochastc ncome n http://www.prnceton.edu/~moll/ HACTproect/HACT_Numercal_Appendx.pdf,

More information

A Local Variational Problem of Second Order for a Class of Optimal Control Problems with Nonsmooth Objective Function

A Local Variational Problem of Second Order for a Class of Optimal Control Problems with Nonsmooth Objective Function A Local Varatonal Problem of Second Order for a Class of Optmal Control Problems wth Nonsmooth Objectve Functon Alexander P. Afanasev Insttute for Informaton Transmsson Problems, Russan Academy of Scences,

More information

PHYS 705: Classical Mechanics. Calculus of Variations II

PHYS 705: Classical Mechanics. Calculus of Variations II 1 PHYS 705: Classcal Mechancs Calculus of Varatons II 2 Calculus of Varatons: Generalzaton (no constrant yet) Suppose now that F depends on several dependent varables : We need to fnd such that has a statonary

More information

Global Sensitivity. Tuesday 20 th February, 2018

Global Sensitivity. Tuesday 20 th February, 2018 Global Senstvty Tuesday 2 th February, 28 ) Local Senstvty Most senstvty analyses [] are based on local estmates of senstvty, typcally by expandng the response n a Taylor seres about some specfc values

More information

Stanford University CS359G: Graph Partitioning and Expanders Handout 4 Luca Trevisan January 13, 2011

Stanford University CS359G: Graph Partitioning and Expanders Handout 4 Luca Trevisan January 13, 2011 Stanford Unversty CS359G: Graph Parttonng and Expanders Handout 4 Luca Trevsan January 3, 0 Lecture 4 In whch we prove the dffcult drecton of Cheeger s nequalty. As n the past lectures, consder an undrected

More information

Outline. Communication. Bellman Ford Algorithm. Bellman Ford Example. Bellman Ford Shortest Path [1]

Outline. Communication. Bellman Ford Algorithm. Bellman Ford Example. Bellman Ford Shortest Path [1] DYNAMIC SHORTEST PATH SEARCH AND SYNCHRONIZED TASK SWITCHING Jay Wagenpfel, Adran Trachte 2 Outlne Shortest Communcaton Path Searchng Bellmann Ford algorthm Algorthm for dynamc case Modfcatons to our algorthm

More information

6.854J / J Advanced Algorithms Fall 2008

6.854J / J Advanced Algorithms Fall 2008 MIT OpenCourseWare http://ocw.mt.edu 6.854J / 18.415J Advanced Algorthms Fall 2008 For nformaton about ctng these materals or our Terms of Use, vst: http://ocw.mt.edu/terms. 18.415/6.854 Advanced Algorthms

More information

Appendix B. The Finite Difference Scheme

Appendix B. The Finite Difference Scheme 140 APPENDIXES Appendx B. The Fnte Dfference Scheme In ths appendx we present numercal technques whch are used to approxmate solutons of system 3.1 3.3. A comprehensve treatment of theoretcal and mplementaton

More information

12. The Hamilton-Jacobi Equation Michael Fowler

12. The Hamilton-Jacobi Equation Michael Fowler 1. The Hamlton-Jacob Equaton Mchael Fowler Back to Confguraton Space We ve establshed that the acton, regarded as a functon of ts coordnate endponts and tme, satsfes ( ) ( ) S q, t / t+ H qpt,, = 0, and

More information

Vector Norms. Chapter 7 Iterative Techniques in Matrix Algebra. Cauchy-Bunyakovsky-Schwarz Inequality for Sums. Distances. Convergence.

Vector Norms. Chapter 7 Iterative Techniques in Matrix Algebra. Cauchy-Bunyakovsky-Schwarz Inequality for Sums. Distances. Convergence. Vector Norms Chapter 7 Iteratve Technques n Matrx Algebra Per-Olof Persson persson@berkeley.edu Department of Mathematcs Unversty of Calforna, Berkeley Math 128B Numercal Analyss Defnton A vector norm

More information

Pricing and Resource Allocation Game Theoretic Models

Pricing and Resource Allocation Game Theoretic Models Prcng and Resource Allocaton Game Theoretc Models Zhy Huang Changbn Lu Q Zhang Computer and Informaton Scence December 8, 2009 Z. Huang, C. Lu, and Q. Zhang (CIS) Game Theoretc Models December 8, 2009

More information

Equilibrium with Complete Markets. Instructor: Dmytro Hryshko

Equilibrium with Complete Markets. Instructor: Dmytro Hryshko Equlbrum wth Complete Markets Instructor: Dmytro Hryshko 1 / 33 Readngs Ljungqvst and Sargent. Recursve Macroeconomc Theory. MIT Press. Chapter 8. 2 / 33 Equlbrum n pure exchange, nfnte horzon economes,

More information

Market structure and Innovation

Market structure and Innovation Market structure and Innovaton Ths presentaton s based on the paper Market structure and Innovaton authored by Glenn C. Loury, publshed n The Quarterly Journal of Economcs, Vol. 93, No.3 ( Aug 1979) I.

More information

10.34 Numerical Methods Applied to Chemical Engineering Fall Homework #3: Systems of Nonlinear Equations and Optimization

10.34 Numerical Methods Applied to Chemical Engineering Fall Homework #3: Systems of Nonlinear Equations and Optimization 10.34 Numercal Methods Appled to Chemcal Engneerng Fall 2015 Homework #3: Systems of Nonlnear Equatons and Optmzaton Problem 1 (30 ponts). A (homogeneous) azeotrope s a composton of a multcomponent mxture

More information

Solutions to exam in SF1811 Optimization, Jan 14, 2015

Solutions to exam in SF1811 Optimization, Jan 14, 2015 Solutons to exam n SF8 Optmzaton, Jan 4, 25 3 3 O------O -4 \ / \ / The network: \/ where all lnks go from left to rght. /\ / \ / \ 6 O------O -5 2 4.(a) Let x = ( x 3, x 4, x 23, x 24 ) T, where the varable

More information

Winter 2008 CS567 Stochastic Linear/Integer Programming Guest Lecturer: Xu, Huan

Winter 2008 CS567 Stochastic Linear/Integer Programming Guest Lecturer: Xu, Huan Wnter 2008 CS567 Stochastc Lnear/Integer Programmng Guest Lecturer: Xu, Huan Class 2: More Modelng Examples 1 Capacty Expanson Capacty expanson models optmal choces of the tmng and levels of nvestments

More information

EEE 241: Linear Systems

EEE 241: Linear Systems EEE : Lnear Systems Summary #: Backpropagaton BACKPROPAGATION The perceptron rule as well as the Wdrow Hoff learnng were desgned to tran sngle layer networks. They suffer from the same dsadvantage: they

More information

Time-Varying Systems and Computations Lecture 6

Time-Varying Systems and Computations Lecture 6 Tme-Varyng Systems and Computatons Lecture 6 Klaus Depold 14. Januar 2014 The Kalman Flter The Kalman estmaton flter attempts to estmate the actual state of an unknown dscrete dynamcal system, gven nosy

More information

LECTURE 9 CANONICAL CORRELATION ANALYSIS

LECTURE 9 CANONICAL CORRELATION ANALYSIS LECURE 9 CANONICAL CORRELAION ANALYSIS Introducton he concept of canoncal correlaton arses when we want to quantfy the assocatons between two sets of varables. For example, suppose that the frst set of

More information

FUZZY GOAL PROGRAMMING VS ORDINARY FUZZY PROGRAMMING APPROACH FOR MULTI OBJECTIVE PROGRAMMING PROBLEM

FUZZY GOAL PROGRAMMING VS ORDINARY FUZZY PROGRAMMING APPROACH FOR MULTI OBJECTIVE PROGRAMMING PROBLEM Internatonal Conference on Ceramcs, Bkaner, Inda Internatonal Journal of Modern Physcs: Conference Seres Vol. 22 (2013) 757 761 World Scentfc Publshng Company DOI: 10.1142/S2010194513010982 FUZZY GOAL

More information

Tracking with Kalman Filter

Tracking with Kalman Filter Trackng wth Kalman Flter Scott T. Acton Vrgna Image and Vdeo Analyss (VIVA), Charles L. Brown Department of Electrcal and Computer Engneerng Department of Bomedcal Engneerng Unversty of Vrgna, Charlottesvlle,

More information

The Feynman path integral

The Feynman path integral The Feynman path ntegral Aprl 3, 205 Hesenberg and Schrödnger pctures The Schrödnger wave functon places the tme dependence of a physcal system n the state, ψ, t, where the state s a vector n Hlbert space

More information

Modeling and Design of Real-Time Pricing Systems Based on Markov Decision Processes

Modeling and Design of Real-Time Pricing Systems Based on Markov Decision Processes Appled Mathematcs, 04, 5, 485-495 Publshed Onlne June 04 n ScRes. http://www.scrp.org/journal/am http://dx.do.org/0.436/am.04.504 Modelng and Desgn of Real-Tme Prcng Systems Based on Markov Decson Processes

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

Canonical transformations

Canonical transformations Canoncal transformatons November 23, 2014 Recall that we have defned a symplectc transformaton to be any lnear transformaton M A B leavng the symplectc form nvarant, Ω AB M A CM B DΩ CD Coordnate transformatons,

More information

Simultaneous Optimization of Berth Allocation, Quay Crane Assignment and Quay Crane Scheduling Problems in Container Terminals

Simultaneous Optimization of Berth Allocation, Quay Crane Assignment and Quay Crane Scheduling Problems in Container Terminals Smultaneous Optmzaton of Berth Allocaton, Quay Crane Assgnment and Quay Crane Schedulng Problems n Contaner Termnals Necat Aras, Yavuz Türkoğulları, Z. Caner Taşkın, Kuban Altınel Abstract In ths work,

More information

Structure and Drive Paul A. Jensen Copyright July 20, 2003

Structure and Drive Paul A. Jensen Copyright July 20, 2003 Structure and Drve Paul A. Jensen Copyrght July 20, 2003 A system s made up of several operatons wth flow passng between them. The structure of the system descrbes the flow paths from nputs to outputs.

More information

Kernel Methods and SVMs Extension

Kernel Methods and SVMs Extension Kernel Methods and SVMs Extenson The purpose of ths document s to revew materal covered n Machne Learnng 1 Supervsed Learnng regardng support vector machnes (SVMs). Ths document also provdes a general

More information

Supplement: Proofs and Technical Details for The Solution Path of the Generalized Lasso

Supplement: Proofs and Technical Details for The Solution Path of the Generalized Lasso Supplement: Proofs and Techncal Detals for The Soluton Path of the Generalzed Lasso Ryan J. Tbshran Jonathan Taylor In ths document we gve supplementary detals to the paper The Soluton Path of the Generalzed

More information

Report on Image warping

Report on Image warping Report on Image warpng Xuan Ne, Dec. 20, 2004 Ths document summarzed the algorthms of our mage warpng soluton for further study, and there s a detaled descrpton about the mplementaton of these algorthms.

More information

Stability analysis for class of switched nonlinear systems

Stability analysis for class of switched nonlinear systems Stablty analyss for class of swtched nonlnear systems The MIT Faculty has made ths artcle openly avalable. Please share how ths access benefts you. Your story matters. Ctaton As Publshed Publsher Shaker,

More information

CHAPTER 5 NUMERICAL EVALUATION OF DYNAMIC RESPONSE

CHAPTER 5 NUMERICAL EVALUATION OF DYNAMIC RESPONSE CHAPTER 5 NUMERICAL EVALUATION OF DYNAMIC RESPONSE Analytcal soluton s usually not possble when exctaton vares arbtrarly wth tme or f the system s nonlnear. Such problems can be solved by numercal tmesteppng

More information

Estimation: Part 2. Chapter GREG estimation

Estimation: Part 2. Chapter GREG estimation Chapter 9 Estmaton: Part 2 9. GREG estmaton In Chapter 8, we have seen that the regresson estmator s an effcent estmator when there s a lnear relatonshp between y and x. In ths chapter, we generalzed the

More information

VQ widely used in coding speech, image, and video

VQ widely used in coding speech, image, and video at Scalar quantzers are specal cases of vector quantzers (VQ): they are constraned to look at one sample at a tme (memoryless) VQ does not have such constrant better RD perfomance expected Source codng

More information

The equation of motion of a dynamical system is given by a set of differential equations. That is (1)

The equation of motion of a dynamical system is given by a set of differential equations. That is (1) Dynamcal Systems Many engneerng and natural systems are dynamcal systems. For example a pendulum s a dynamcal system. State l The state of the dynamcal system specfes t condtons. For a pendulum n the absence

More information

Inexact Alternating Minimization Algorithm for Distributed Optimization with an Application to Distributed MPC

Inexact Alternating Minimization Algorithm for Distributed Optimization with an Application to Distributed MPC Inexact Alternatng Mnmzaton Algorthm for Dstrbuted Optmzaton wth an Applcaton to Dstrbuted MPC Ye Pu, Coln N. Jones and Melane N. Zelnger arxv:608.0043v [math.oc] Aug 206 Abstract In ths paper, we propose

More information

College of Computer & Information Science Fall 2009 Northeastern University 20 October 2009

College of Computer & Information Science Fall 2009 Northeastern University 20 October 2009 College of Computer & Informaton Scence Fall 2009 Northeastern Unversty 20 October 2009 CS7880: Algorthmc Power Tools Scrbe: Jan Wen and Laura Poplawsk Lecture Outlne: Prmal-dual schema Network Desgn:

More information

A New Algorithm for Finding a Fuzzy Optimal. Solution for Fuzzy Transportation Problems

A New Algorithm for Finding a Fuzzy Optimal. Solution for Fuzzy Transportation Problems Appled Mathematcal Scences, Vol. 4, 200, no. 2, 79-90 A New Algorthm for Fndng a Fuzzy Optmal Soluton for Fuzzy Transportaton Problems P. Pandan and G. Nataraan Department of Mathematcs, School of Scence

More information

Appendix for Causal Interaction in Factorial Experiments: Application to Conjoint Analysis

Appendix for Causal Interaction in Factorial Experiments: Application to Conjoint Analysis A Appendx for Causal Interacton n Factoral Experments: Applcaton to Conjont Analyss Mathematcal Appendx: Proofs of Theorems A. Lemmas Below, we descrbe all the lemmas, whch are used to prove the man theorems

More information

On a direct solver for linear least squares problems

On a direct solver for linear least squares problems ISSN 2066-6594 Ann. Acad. Rom. Sc. Ser. Math. Appl. Vol. 8, No. 2/2016 On a drect solver for lnear least squares problems Constantn Popa Abstract The Null Space (NS) algorthm s a drect solver for lnear

More information

One-Sided Adaptation for Infinite-Horizon Linear Quadratic N-person Nonzero-sum Dynamic Games

One-Sided Adaptation for Infinite-Horizon Linear Quadratic N-person Nonzero-sum Dynamic Games Proceedngs of the European Control Conference 7 Kos, Greece, July -5, 7 TuB45 One-Sded Adaptaton for Infnte-Horzon Lnear Quadratc N-person Nonzero-sum Dynamc Games Xaohuan Tan and Jose B Cruz, Jr, Lfe

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

Some modelling aspects for the Matlab implementation of MMA

Some modelling aspects for the Matlab implementation of MMA Some modellng aspects for the Matlab mplementaton of MMA Krster Svanberg krlle@math.kth.se Optmzaton and Systems Theory Department of Mathematcs KTH, SE 10044 Stockholm September 2004 1. Consdered optmzaton

More information

PROBLEM SET 7 GENERAL EQUILIBRIUM

PROBLEM SET 7 GENERAL EQUILIBRIUM PROBLEM SET 7 GENERAL EQUILIBRIUM Queston a Defnton: An Arrow-Debreu Compettve Equlbrum s a vector of prces {p t } and allocatons {c t, c 2 t } whch satsfes ( Gven {p t }, c t maxmzes βt ln c t subject

More information

Singular Value Decomposition: Theory and Applications

Singular Value Decomposition: Theory and Applications Sngular Value Decomposton: Theory and Applcatons Danel Khashab Sprng 2015 Last Update: March 2, 2015 1 Introducton A = UDV where columns of U and V are orthonormal and matrx D s dagonal wth postve real

More information

Deriving the X-Z Identity from Auxiliary Space Method

Deriving the X-Z Identity from Auxiliary Space Method Dervng the X-Z Identty from Auxlary Space Method Long Chen Department of Mathematcs, Unversty of Calforna at Irvne, Irvne, CA 92697 chenlong@math.uc.edu 1 Iteratve Methods In ths paper we dscuss teratve

More information

CSci 6974 and ECSE 6966 Math. Tech. for Vision, Graphics and Robotics Lecture 21, April 17, 2006 Estimating A Plane Homography

CSci 6974 and ECSE 6966 Math. Tech. for Vision, Graphics and Robotics Lecture 21, April 17, 2006 Estimating A Plane Homography CSc 6974 and ECSE 6966 Math. Tech. for Vson, Graphcs and Robotcs Lecture 21, Aprl 17, 2006 Estmatng A Plane Homography Overvew We contnue wth a dscusson of the major ssues, usng estmaton of plane projectve

More information

NON-CENTRAL 7-POINT FORMULA IN THE METHOD OF LINES FOR PARABOLIC AND BURGERS' EQUATIONS

NON-CENTRAL 7-POINT FORMULA IN THE METHOD OF LINES FOR PARABOLIC AND BURGERS' EQUATIONS IJRRAS 8 (3 September 011 www.arpapress.com/volumes/vol8issue3/ijrras_8_3_08.pdf NON-CENTRAL 7-POINT FORMULA IN THE METHOD OF LINES FOR PARABOLIC AND BURGERS' EQUATIONS H.O. Bakodah Dept. of Mathematc

More information

Inductance Calculation for Conductors of Arbitrary Shape

Inductance Calculation for Conductors of Arbitrary Shape CRYO/02/028 Aprl 5, 2002 Inductance Calculaton for Conductors of Arbtrary Shape L. Bottura Dstrbuton: Internal Summary In ths note we descrbe a method for the numercal calculaton of nductances among conductors

More information

The Two-scale Finite Element Errors Analysis for One Class of Thermoelastic Problem in Periodic Composites

The Two-scale Finite Element Errors Analysis for One Class of Thermoelastic Problem in Periodic Composites 7 Asa-Pacfc Engneerng Technology Conference (APETC 7) ISBN: 978--6595-443- The Two-scale Fnte Element Errors Analyss for One Class of Thermoelastc Problem n Perodc Compostes Xaoun Deng Mngxang Deng ABSTRACT

More information

Chapter 11: Simple Linear Regression and Correlation

Chapter 11: Simple Linear Regression and Correlation Chapter 11: Smple Lnear Regresson and Correlaton 11-1 Emprcal Models 11-2 Smple Lnear Regresson 11-3 Propertes of the Least Squares Estmators 11-4 Hypothess Test n Smple Lnear Regresson 11-4.1 Use of t-tests

More information

LINEAR REGRESSION ANALYSIS. MODULE IX Lecture Multicollinearity

LINEAR REGRESSION ANALYSIS. MODULE IX Lecture Multicollinearity LINEAR REGRESSION ANALYSIS MODULE IX Lecture - 30 Multcollnearty Dr. Shalabh Department of Mathematcs and Statstcs Indan Insttute of Technology Kanpur 2 Remedes for multcollnearty Varous technques have

More information