Expectations or Guarantees? I Want It All! A Crossroad between Games and MDPs
|
|
- Vincent Paul
- 5 years ago
- Views:
Transcription
1 Expectations or Guarantees? I Want It All! A Crossroad between Games and MDPs V. Bruyère (UMONS) E. Filiot (ULB) M. Randour (UMONS-ULB) J.-F. Raskin (ULB) Grenoble SR 04 - nd International Workshop on Strategic Reasoning
2 The talk in two slides (/) Verification and synthesis: a reactive system to control, an interacting environment, a specification to enforce. Focus on quantitative properties. Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin / 6
3 The talk in two slides (/) Verification and synthesis: a reactive system to control, an interacting environment, a specification to enforce. Focus on quantitative properties. Several ways to look at the interactions, and in particular, the nature of the environment. Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin / 6
4 The talk in two slides (/) Games antagonistic adversary guarantees on worst-case MDPs stochastic adversary optimize expected value Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin / 6
5 The talk in two slides (/) Games antagonistic adversary guarantees on worst-case MDPs stochastic adversary optimize expected value BWC synthesis ensure both Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin / 6
6 The talk in two slides (/) Games antagonistic adversary guarantees on worst-case MDPs stochastic adversary optimize expected value BWC synthesis ensure both Mean-Payoff Studied value functions Shortest Path Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin / 6
7 Advertisement Featured in STACS 4 [BFRR4] Full paper available on arxiv: abs/ Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 3 / 6
8 Context BWC Synthesis 3 Mean-Payoff 4 Shortest Path Conclusion Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 4 / 6
9 Context BWC Synthesis 3 Mean-Payoff 4 Shortest Path Conclusion Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin / 6
10 Quantitative games on graphs 4 7 Graph G = (S, E, w) with w : E Z Two-player game G = (G, S, S ) P states = P states = Plays have values f : Plays(G) R {, } Players follow strategies λ i : Prefs i (G) D(S) Finite memory stochastic output Moore machine M(λ i ) = (Mem, m 0, α u, α n ) Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 6 / 6
11 Quantitative games on graphs 4 7 Graph G = (S, E, w) with w : E Z Two-player game G = (G, S, S ) P states = P states = Plays have values f : Plays(G) R {, } Players follow strategies λ i : Prefs i (G) D(S) Finite memory stochastic output Moore machine M(λ i ) = (Mem, m 0, α u, α n ) Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 6 / 6
12 Quantitative games on graphs 4 7 Graph G = (S, E, w) with w : E Z Two-player game G = (G, S, S ) P states = P states = Plays have values f : Plays(G) R {, } Players follow strategies λ i : Prefs i (G) D(S) Finite memory stochastic output Moore machine M(λ i ) = (Mem, m 0, α u, α n ) Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 6 / 6
13 Quantitative games on graphs 4 7 Graph G = (S, E, w) with w : E Z Two-player game G = (G, S, S ) P states = P states = Plays have values f : Plays(G) R {, } Players follow strategies λ i : Prefs i (G) D(S) Finite memory stochastic output Moore machine M(λ i ) = (Mem, m 0, α u, α n ) Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 6 / 6
14 Quantitative games on graphs 4 7 Graph G = (S, E, w) with w : E Z Two-player game G = (G, S, S ) P states = P states = Plays have values f : Plays(G) R {, } Players follow strategies λ i : Prefs i (G) D(S) Finite memory stochastic output Moore machine M(λ i ) = (Mem, m 0, α u, α n ) Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 6 / 6
15 Quantitative games on graphs 4 7 Graph G = (S, E, w) with w : E Z Two-player game G = (G, S, S ) P states = P states = Plays have values f : Plays(G) R {, } Players follow strategies λ i : Prefs i (G) D(S) Finite memory stochastic output Moore machine M(λ i ) = (Mem, m 0, α u, α n ) Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 6 / 6
16 Quantitative games on graphs 7 4 Then, (,, ) ω Graph G = (S, E, w) with w : E Z Two-player game G = (G, S, S ) P states = P states = Plays have values f : Plays(G) R {, } Players follow strategies λ i : Prefs i (G) D(S) Finite memory stochastic output Moore machine M(λ i ) = (Mem, m 0, α u, α n ) Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 6 / 6
17 Markov decision processes 4 7 MDP P = (G, S, S, ) with : S D(S) P states = stochastic states = MDP = game + strategy of P P = G[λ ] Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 7 / 6
18 Markov chains MC M = (G, δ) with δ : S D(S) 4 MC = MDP + strategy of P = game + both strategies M = P[λ ] = G[λ, λ ] Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 8 / 6
19 Markov chains MC M = (G, δ) with δ : S D(S) MC = MDP + strategy of P = game + both strategies M = P[λ ] = G[λ, λ ] Event A Plays(G) probability P M s init (A) Measurable f : Plays(G) R {, } expected value E M s init (f ) Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 8 / 6
20 Classical interpretations System trying to ensure a specification = P whatever the actions of its environment Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 9 / 6
21 Classical interpretations System trying to ensure a specification = P whatever the actions of its environment The environment can be seen as antagonistic two-player game, worst-case threshold problem for µ Q? λ Λ, λ Λ, π Outs G (s init, λ, λ ), f (π) µ Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 9 / 6
22 Classical interpretations System trying to ensure a specification = P whatever the actions of its environment The environment can be seen as antagonistic two-player game, worst-case threshold problem for µ Q? λ Λ, λ Λ, π Outs G (s init, λ, λ ), f (π) µ fully stochastic MDP, expected value threshold problem for ν Q? λ Λ, E P[λ ] s init (f ) ν Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 9 / 6
23 Context BWC Synthesis 3 Mean-Payoff 4 Shortest Path Conclusion Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 0 / 6
24 What if you want both? In practice, we want both nice expected performance in the everyday situation, strict (but relaxed) performance guarantees even in the event of very bad circumstances. Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin / 6
25 Example: going to work back home delay 0 train station 9 0 home departs 3 wait 4 light 0 0 car traffic 7 0 medium 30 0 heavy 70 bicycle 4 Weights = minutes Goal: minimize our expected time to reach work But, important meeting in one hour! Requires strict guarantees on the worst-case reaching time. waiting room work Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin / 6
26 Example: going to work back home delay 0 train station 9 0 home departs 3 wait 4 light 0 0 car traffic 7 0 medium 30 0 heavy 70 bicycle 4 Optimal expectation strategy: take the car. E = 33, WC = 7 > 60. Optimal worst-case strategy: bicycle. E = WC = 4 < 60. waiting room work Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin / 6
27 Example: going to work back home delay 0 train station waiting room 9 0 home departs 3 wait 4 light 0 0 car traffic 7 0 medium 30 work 0 heavy 70 bicycle 4 Optimal expectation strategy: take the car. E = 33, WC = 7 > 60. Optimal worst-case strategy: bicycle. E = WC = 4 < 60. Sample BWC strategy: try train up to 3 delays then switch to bicycle. E 37.6, WC = 9 < 60. Optimal E under WC constraint Uses finite memory Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin / 6
28 Beyond worst-case synthesis Formal definition Given a game G = (G, S, S ), with G = (S, E, w) its underlying graph, an initial state s init S, a finite-memory stochastic model λ stoch Λ F of the adversary, represented by a stochastic Moore machine, a measurable value function f : Plays(G) R {, }, and two rational thresholds µ, ν Q, the beyond worst-case (BWC) problem asks to decide if P has a finite-memory strategy λ Λ F such that { λ Λ, π Outs G (s init, λ, λ ), f (π) > µ () E G[λ,λ stoch ] s init (f ) > ν () and the BWC synthesis problem asks to synthesize such a strategy if one exists. Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 3 / 6
29 Beyond worst-case synthesis Formal definition Given a game G = (G, S, S ), with G = (S, E, w) its underlying graph, an initial state s init S, a finite-memory stochastic model λ stoch Λ F of the adversary, represented by a stochastic Moore machine, a measurable value function f : Plays(G) R {, }, and two rational thresholds µ, ν Q, the beyond worst-case (BWC) problem asks to decide if P has a finite-memory strategy λ Λ F such that { λ Λ, π Outs G (s init, λ, λ ), f (π) > µ () E G[λ,λ stoch ] s init (f ) > ν () and the BWC synthesis problem asks to synthesize such a strategy if one exists. Notice the highlighted parts! Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 3 / 6
30 Related work Common philosophy: avoiding outlier outcomes Our strategies are strongly risk averse avoid risk at all costs and optimize among safe strategies Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 4 / 6
31 Related work Common philosophy: avoiding outlier outcomes Our strategies are strongly risk averse avoid risk at all costs and optimize among safe strategies Other notions of risk ensure low probability of risked behavior [WL99, FKR9] without worst-case guarantee without good expectation Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 4 / 6
32 Related work Common philosophy: avoiding outlier outcomes Our strategies are strongly risk averse avoid risk at all costs and optimize among safe strategies Other notions of risk ensure low probability of risked behavior [WL99, FKR9] without worst-case guarantee without good expectation 3 Trade-off between expectation and variance [BCFK3, MT] statistical measure of the stability of the performance no strict guarantee on individual outcomes Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 4 / 6
33 Context BWC Synthesis 3 Mean-Payoff 4 Shortest Path Conclusion Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin / 6
34 Mean-payoff value function MP(π) = lim inf n [ i=n n w ( (s i, s i+ ) )] i=0 Sample play π =,, 4,, (,, ) ω MP(π) = 3 long-run average weight prefix-independent Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 6 / 6
35 Mean-payoff value function MP(π) = lim inf n [ i=n n w ( (s i, s i+ ) )] i=0 Sample play π =,, 4,, (,, ) ω MP(π) = 3 long-run average weight prefix-independent worst-case expected value BWC complexity NP conp P NP conp memory memoryless memoryless pseudo-polynomial [LL69, EM79, ZP96, Jur98, GS09, Put94, FV97] Additional modeling power for free! Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 6 / 6
36 Philosophy of the algorithm Classical worst-case and expected value results and algorithms as nuts and bolts Screw them together in an adequate way Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 7 / 6
37 Philosophy of the algorithm Classical worst-case and expected value results and algorithms as nuts and bolts Screw them together in an adequate way Three key ideas To characterize the expected value, look at end-components (ECs) Winning ECs vs. losing ECs: the latter must be avoided to preserve the worst-case requirement! 3 Inside a WEC, we have an interesting way to play... Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 7 / 6
38 Philosophy of the algorithm Classical worst-case and expected value results and algorithms as nuts and bolts Screw them together in an adequate way Three key ideas To characterize the expected value, look at end-components (ECs) Winning ECs vs. losing ECs: the latter must be avoided to preserve the worst-case requirement! 3 Inside a WEC, we have an interesting way to play... = Let s focus on an ideal case Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 7 / 6
39 An ideal situation s s 6 s Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 8 / 6
40 An ideal situation s s 6 s Game interpretation Worst-case threshold is µ = 0 All states are winning: memoryless optimal worst-case strategy λ wc Λ PM (G), ensuring µ = > 0 Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 8 / 6
41 An ideal situation s s 6 s Game interpretation Worst-case threshold is µ = 0 All states are winning: memoryless optimal worst-case strategy λ wc Λ PM (G), ensuring µ = > 0 MDP interpretation Memoryless optimal expected value strategy λ e ΛPM (P) achieves ν = Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 8 / 6
42 A cornerstone of our approach s s 6 s BWC problem: what kind of threholds (0, ν) can we achieve? Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 8 / 6
43 A cornerstone of our approach s s 6 s BWC problem: what kind of threholds (0, ν) can we achieve? Key result For all ε > 0, there exists a finite-memory strategy of P that satisfies the BWC problem for the thresholds pair (0, ν ε). We can be arbitrarily close to the optimal expectation while ensuring the worst-case! Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 8 / 6
44 Combined strategy s s 6 s Outcomes of the form > 0 > 0 0 compensate K steps L steps > 0 0 compensate WC > 0 E =?? Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 8 / 6
45 Combined strategy s s 6 s Outcomes of the form > 0 > 0 0 K steps L steps What we want compensate > 0 0 compensate K, L WC > 0 E =?? E = ν = Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 8 / 6
46 Combined strategy: crux of the proof Precise reasoning on convergence rates using involved techniques When K grows, L needs to grow linearly to ensure WC Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 9 / 6
47 Combined strategy: crux of the proof Precise reasoning on convergence rates using involved techniques When K grows, L needs to grow linearly to ensure WC When K grows, P( ) 0 and it decreases exponentially fast application of Chernoff bounds and Hoeffding s inequality for Markov chains [Tra09, GO0] Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 9 / 6
48 Combined strategy: crux of the proof Precise reasoning on convergence rates using involved techniques When K grows, L needs to grow linearly to ensure WC When K grows, P( ) 0 and it decreases exponentially fast application of Chernoff bounds and Hoeffding s inequality for Markov chains [Tra09, GO0] Overall we are good: WC > 0 and E > ν ε for sufficiently large K, L. Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 9 / 6
49 Context BWC Synthesis 3 Mean-Payoff 4 Shortest Path Conclusion Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 0 / 6
50 Shortest path - truncated sum Assume strictly positive integer weights, w : E N 0 Let T S be a target set that P wants to reach with a path of bounded value (cf. introductory example) inequalities are reversed, ν < µ TS T (π = s 0 s s... ) = n i=0 w((s i, s i+ )), with n the first index such that s n T, and TS T (π) = if n, s n T Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin / 6
51 Shortest path - truncated sum Assume strictly positive integer weights, w : E N 0 Let T S be a target set that P wants to reach with a path of bounded value (cf. introductory example) inequalities are reversed, ν < µ TS T (π = s 0 s s... ) = n i=0 w((s i, s i+ )), with n the first index such that s n T, and TS T (π) = if n, s n T worst-case expected value BWC complexity P P pseudo-poly. / NP-hard memory memoryless memoryless pseudo-poly. [BT9, da99] Problem inherently harder than worst-case and expectation. NP-hardness by K th largest subset problem [JK78, GJ79] Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin / 6
52 Key difference with MP case Useful observation The set of all worst-case winning strategies for the shortest path can be represented through a finite game. Sequential approach solving the BWC problem: represent all WC winning strategies, optimize the expected value within those strategies. Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin / 6
53 Pseudo-polynomial algorithm: sketch s s s 3 Start from G = (G, S, S ), G = (S, E, w), T = {s 3 }, M(λ stoch ), µ = 8, and ν Q Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 3 / 6
54 Pseudo-polynomial algorithm: sketch s s s 3 Start from G = (G, S, S ), G = (S, E, w), T = {s 3 }, M(λ stoch ), µ = 8, and ν Q Build G by unfolding G, tracking the current sum up to the worst-case threshold µ, and integrating it in the states of G. Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 3 / 6
55 Pseudo-polynomial algorithm: sketch s s s 3 s, 0 s, s3, Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 3 / 6
56 Pseudo-polynomial algorithm: sketch s s s 3 s, 0 s, s, s3, s3, Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 3 / 6
57 Pseudo-polynomial algorithm: sketch s s s 3 s, 0 s, s, s, 3 s3, s3, s3, 7 Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 3 / 6
58 Pseudo-polynomial algorithm: sketch s s s 3 s, 0 s, s, s, 3 s, 4 s3, s3, s3, 7 s3, 4 Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 3 / 6
59 Pseudo-polynomial algorithm: sketch s s s 3 s, 0 s, s, s, 3 s, 4 s, s3, s3, s3, 7 s3, 4 s3, Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 3 / 6
60 Pseudo-polynomial algorithm: sketch s s s 3 s, 0 s, s, s, 3 s, 4 s, s, 6 s3, s3, s3, 7 s3, 4 s3, s3, 6 Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 3 / 6
61 Pseudo-polynomial algorithm: sketch s s s 3 s, 0 s, s, s, 3 s, 4 s, s, 6 s, 7 s, s3, s3, s3, 7 s3, 4 s3, s3, 6 Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 3 / 6
62 Pseudo-polynomial algorithm: sketch 3 Compute R, the attractor of T with cost < µ = 8 4 Consider G µ = G R s, 0 s, s, s, 3 s, 4 s, s, 6 s, 7 s, s3, s3, s3, 7 s3, 4 s3, s3, 6 Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 3 / 6
63 Pseudo-polynomial algorithm: sketch 3 Compute R, the attractor of T with cost < µ = 8 4 Consider G µ = G R s, 0 s, s, s3, s3, s3, 7 Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 3 / 6
64 Pseudo-polynomial algorithm: sketch Consider P = G µ M(λ stoch ) 6 Compute memoryless optimal expectation strategy 7 If ν < ν, answer Yes, otherwise answer No s, 0 s, s, Here, ν = 9/ s3, s3, s3, 7 Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 3 / 6
65 Context BWC Synthesis 3 Mean-Payoff 4 Shortest Path Conclusion Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 4 / 6
66 In a nutshell BWC framework combines worst-case and expected value requirements a natural wish in many practical applications few existing theoretical support Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin / 6
67 In a nutshell BWC framework combines worst-case and expected value requirements a natural wish in many practical applications few existing theoretical support Mean-payoff: additional modeling power for no complexity cost (decision-wise) Shortest path: harder than the worst-case, pseudo-polynomial with NP-hardness result Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin / 6
68 In a nutshell BWC framework combines worst-case and expected value requirements a natural wish in many practical applications few existing theoretical support Mean-payoff: additional modeling power for no complexity cost (decision-wise) Shortest path: harder than the worst-case, pseudo-polynomial with NP-hardness result In both cases, pseudo-polynomial memory is both sufficient and necessary but strategies have natural representations based on states of the game and simple integer counters Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin / 6
69 Beyond BWC synthesis? Possible future works include study of other quantitative objectives, extension of our results to more general settings (multi-dimension [CDHR0, CRR], decidable classes of games with imperfect information [DDG + 0], etc), application of the BWC problem to various practical cases. Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 6 / 6
70 Beyond BWC synthesis? Possible future works include study of other quantitative objectives, extension of our results to more general settings (multi-dimension [CDHR0, CRR], decidable classes of games with imperfect information [DDG + 0], etc), application of the BWC problem to various practical cases. Thanks! Do not hesitate to discuss with us! Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 6 / 6
71 References I T. Brázdil, K. Chatterjee, V. Forejt, and A. Kucera. Trading performance for stability in Markov decision processes. In Proc. of LICS, pages IEEE Computer Society, 03. V. Bruyère, E. Filiot, M. Randour, and J.-F. Raskin. Meet your expectations with guarantees: beyond worst-case synthesis in quantitative games. In Proc. of STACS, LIPIcs, pages Schloss Dagstuhl - LZI, 04. D.P. Bertsekas and J.N. Tsitsiklis. An analysis of stochastic shortest path problems. Mathematics of Operations Research, 6:80 9, 99. K. Chatterjee, L. Doyen, T.A. Henzinger, and J.-F. Raskin. Generalized mean-payoff and energy games. In Proc. of FSTTCS, LIPIcs 8, pages 0 6. Schloss Dagstuhl - LZI, 00. K. Chatterjee, L. Doyen, M. Randour, and J.-F. Raskin. Looking at mean-payoff and total-payoff through windows. In Proc. of ATVA, LNCS 87, pages 8 3. Springer, 03. K. Chatterjee, M. Randour, and J.-F. Raskin. Strategy synthesis for multi-dimensional quantitative objectives. In Proc. of CONCUR, LNCS 744, pages 3. Springer, 0. L. de Alfaro. Computing minimum and maximum reachability times in probabilistic systems. In Proc. of CONCUR, LNCS 664, pages Springer, 999. Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 7 / 6
72 References II A. Degorre, L. Doyen, R. Gentilini, J.-F. Raskin, and S. Torunczyk. Energy and mean-payoff games with imperfect information. In Proc. of CSL, LNCS 647, pages Springer, 00. A. Ehrenfeucht and J. Mycielski. Positional strategies for mean payoff games. Int. Journal of Game Theory, 8():09 3, 979. J.A. Filar, D. Krass, and K.W. Ross. Percentile performance criteria for limiting average Markov decision processes. Transactions on Automatic Control, pages 0, 99. J. Filar and K. Vrieze. Competitive Markov decision processes. Springer, 997. M.R. Garey and D.S. Johnson. Computers and intractability: a guide to the Theory of NP-Completeness. Freeman New York, 979. P.W. Glynn and D. Ormoneit. Hoeffding s inequality for uniformly ergodic Markov chains. Statistics & Probability Letters, 6():43 46, 00. T. Gawlitza and H. Seidl. Games through nested fixpoints. In Proc. of CAV, LNCS 643, pages Springer, 009. Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 8 / 6
73 References III D.B. Johnson and S.D. Kashdan. Lower bounds for selection in X + Y and other multisets. Journal of the ACM, (4):6 70, 978. M. Jurdziński. Deciding the winner in parity games is in UP co-up. Inf. Process. Lett., 68(3):9 4, 998. T.M. Liggett and S.A. Lippman. Stochastic games with perfect information and time average payoff. Siam Review, (4): , 969. S. Mannor and J.N. Tsitsiklis. Mean-variance optimization in Markov decision processes. In Proc. of ICML, pages Omnipress, 0. M.L. Puterman. Markov decision processes: discrete stochastic dynamic programming. John Wiley & Sons, Inc., New York, NY, USA, st edition, 994. M. Tracol. Fast convergence to state-action frequency polytopes for MDPs. Oper. Res. Lett., 37():3 6, 009. C. Wu and Y. Lin. Minimizing risk models in Markov decision processes with policies depending on target values. Journal of Mathematical Analysis and Applications, 3():47 67, 999. Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 9 / 6
74 References IV U. Zwick and M. Paterson. The complexity of mean payoff games on graphs. Theoretical Computer Science, 8:343 39, 996. Beyond Worst-Case Synthesis Bruyère, Filiot, Randour, Raskin 30 / 6
Looking at Mean-Payoff and Total-Payoff through Windows
Looking at Mean-Payoff and Total-Payoff through Windows Krishnendu Chatterjee 1,, Laurent Doyen 2, Mickael Randour 3,, and Jean-François Raskin 4, 1 IST Austria (Institute of Science and Technology Austria)
More informationMarkov Decision Processes with Multiple Long-run Average Objectives
Markov Decision Processes with Multiple Long-run Average Objectives Krishnendu Chatterjee UC Berkeley c krish@eecs.berkeley.edu Abstract. We consider Markov decision processes (MDPs) with multiple long-run
More informationRandomness for Free. 1 Introduction. Krishnendu Chatterjee 1, Laurent Doyen 2, Hugo Gimbert 3, and Thomas A. Henzinger 1
Randomness for Free Krishnendu Chatterjee 1, Laurent Doyen 2, Hugo Gimbert 3, and Thomas A. Henzinger 1 1 IST Austria (Institute of Science and Technology Austria) 2 LSV, ENS Cachan & CNRS, France 3 LaBri
More informationAverage-Energy Games
Average-Energy Games Patricia Bouyer 1 Kim G. Larsen 2 Simon Laursen 2 Nicolas Markey 1 Mickael Randour 1 1 LSV - CNRS & ENS Cachan 2 Aalborg University 18.05.2015-5th Cassting Meeting - ULB General context:
More informationAverage-Energy Games
Average-Energy Games Patricia Bouyer 1 Nicolas Markey 1 Mickael Randour 1 Kim G. Larsen 2 Simon Laursen 2 1 LSV - CNRS & ENS Cachan 2 Aalborg University 09.06.2015 - Séminaire annuel du LSV - Dourdan General
More informationarxiv: v4 [cs.lo] 2 Oct 2010
Generalized Mean-payoff and Energy Games Krishnendu Chatterjee 1, Laurent Doyen 2, Thomas A. Henzinger 1, and Jean-François Raskin 3 1 IST Austria (Institute of Science and Technology Austria) 2 LSV, ENS
More informationLooking at Mean-Payoff and Total-Payoff through Windows
Looking at Mean-Payoff and Total-Payoff through Windows Krishnendu Chatterjee a,1, Laurent Doyen b, Mickael Randour b,c,2, Jean-François Raskin d,3 a IST Austria (Institute of Science and Technology Austria)
More informationA Survey of Partial-Observation Stochastic Parity Games
Noname manuscript No. (will be inserted by the editor) A Survey of Partial-Observation Stochastic Parity Games Krishnendu Chatterjee Laurent Doyen Thomas A. Henzinger the date of receipt and acceptance
More informationarxiv: v1 [cs.lo] 17 Dec 2015
Non-Zero Sum Games for Reactive Synthesis Romain Brenguier 1, Lorenzo Clemente 2, Paul Hunter 3, Guillermo A. Pérez 3, Mickael Randour 3, Jean-François Raskin 3, Ocan Sankur 4, Mathieu Sassolas 5 arxiv:1512.05568v1
More informationA Note on the Approximation of Mean-Payoff Games
A Note on the Approximation of Mean-Payoff Games Raffaella Gentilini 1 1 University of Perugia, Italy Abstract. We consider the problem of designing approximation schemes for the values of mean-payoff
More informationEPTCS 146. Proceedings of the 2nd International Workshop on Strategic Reasoning. Grenoble, France, April 5-6, 2014
EPTCS 146 Proceedings of the 2nd International Workshop on Strategic Reasoning Grenoble, France, April 5-6, 2014 Edited by: Fabio Mogavero, Aniello Murano and Moshe Y. Vardi Published: 1st April 2014 DOI:
More informationEnergy and Mean-Payoff Games with Imperfect Information
Energy and Mean-Payoff Games with Imperfect Information Aldric Degorre 1, Laurent Doyen 2, Raffaella Gentilini 3, Jean-François Raskin 1, and Szymon Toruńczyk 2 1 Université Libre de Bruxelles (ULB), Belgium
More informationPerfect Half Space Games
Perfect Half Space Games Thomas Colcombet, Marcin Jurdziński, Ranko Lazić, and Sylvain Schmitz LSV, ENS Paris-Saclay & CNRS & Inria LICS 217, June 23rd, 217 1/1 Energy Parity Games Extended Energy Games
More informationThe Complexity of Solving Stochastic Games on Graphs
The Complexity of Solving Stochastic Games on Graphs Daniel Andersson and Peter Bro Miltersen Department of Computer Science, Aarhus University, Denmark Abstract. We consider some well-known families of
More informationStochastic Games with Time The value Min strategies Max strategies Determinacy Finite-state games Cont.-time Markov chains
Games with Time Finite-state Masaryk University Brno GASICS 00 /39 Outline Finite-state stochastic processes. Games over event-driven stochastic processes. Strategies,, determinacy. Existing results for
More informationApproximate Determinization of Quantitative Automata
Approximate Determinization of Quantitative Automata Udi Boker,2 and Thomas A. Henzinger 2 Hebrew University of Jerusalem 2 IST Austria Abstract Quantitative automata are nondeterministic finite automata
More informationTheoretical Computer Science
Theoretical Computer Science 458 (2012) 49 60 Contents lists available at SciVerse ScienceDirect Theoretical Computer Science journal homepage: www.elsevier.com/locate/tcs Energy parity games Krishnendu
More informationSolving Partial-Information Stochastic Parity Games
Solving Partial-Information Stochastic Parity ames Sumit Nain and Moshe Y. Vardi Department of Computer Science, Rice University, Houston, Texas, 77005 Email: {nain,vardi}@cs.rice.edu Abstract We study
More informationMean-Payoff Games and the Max-Atom Problem
Mean-Payoff Games and the Max-Atom Problem Albert Atserias Universitat Politècnica de Catalunya Barcelona, Spain Elitza Maneva Universitat Politècnica de Catalunya Barcelona, Spain February 3, 200 Abstract
More informationFaster Pseudopolynomial Algorithms for Mean-Payoff Games
Faster Pseudopolynomial Algorithms for Mean-Payoff Games 1 Faster Pseudopolynomial Algorithms for Mean-Payoff Games L. Doyen, R. Gentilini, and J.-F. Raskin Univ. Libre de Bruxelles Faster Pseudopolynomial
More informationThe priority promotion approach to parity games
The priority promotion approach to parity games Massimo Benerecetti 1, Daniele Dell Erba 1, and Fabio Mogavero 2 1 Università degli Studi di Napoli Federico II 2 Università degli Studi di Verona Abstract.
More informationA Survey of Stochastic ω-regular Games
A Survey of Stochastic ω-regular Games Krishnendu Chatterjee Thomas A. Henzinger EECS, University of California, Berkeley, USA Computer and Communication Sciences, EPFL, Switzerland {c krish,tah}@eecs.berkeley.edu
More informationThe Complexity of Admissibility in Omega-Regular Games
The Complexity of Admissibility in Omega-Regular Games Romain Brenguier Département d informatique, Université Libre de Bruxelles (U.L.B.), Belgium romain.brenguier@ulb.ac.be Jean-François Raskin Département
More informationAlgorithmic Game Theory and Applications. Lecture 15: a brief taster of Markov Decision Processes and Stochastic Games
Algorithmic Game Theory and Applications Lecture 15: a brief taster of Markov Decision Processes and Stochastic Games Kousha Etessami warning 1 The subjects we will touch on today are so interesting and
More informationThe Complexity of Nash Equilibria in Simple Stochastic Multiplayer Games *
The Complexity of Nash Equilibria in Simple Stochastic Multiplayer Games * Michael Ummels 1 and Dominik Wojtczak 2,3 1 RWTH Aachen University, Germany E-Mail: ummels@logic.rwth-aachen.de 2 CWI, Amsterdam,
More informationMeasuring Permissivity in Finite Games
Measuring Permissivity in Finite Games Patricia Bouyer, Marie Duflot, Nicolas Markey, and Gabriel Renault 3 LSV, CNRS & ENS Cachan, France {bouyer,markey}@lsv.ens-cachan.fr LACL, Université Paris, France
More informationAdmissible Strategies for Synthesizing Systems
Admissible Strategies for Synthesizing Systems Ocan Sankur Univ Rennes, Inria, CNRS, IRISA, Rennes Joint with Romain Brenguier (DiffBlue), Guillermo Pérez (Antwerp), and Jean-François Raskin (ULB) (Multiplayer)
More informationPlaying Stochastic Games Precisely
Playing Stochastic Games Precisely Taolue Chen 1, Vojtěch Forejt 1, Marta Kwiatkowska 1, Aistis Simaitis 1, Ashutosh Trivedi 2, and Michael Ummels 3 1 Department of Computer Science, University of Oxford,
More informationDeciding the Winner in Parity Games Is in UP co-up
c 1998 Elsevier Science B.V. Marcin Jurdziński, Deciding the Winner in Parity Games Is in UP co-up. Information Processing Letters 68(3): 119 124, November 1998. Personal use of this material for research/educational
More informationQuantitative Languages
Quantitative Languages Krishnendu Chatterjee 1, Laurent Doyen 2, and Thomas A. Henzinger 2 1 University of California, Santa Cruz 2 EPFL, Lausanne, Switzerland Abstract. Quantitative generalizations of
More informationPerfect-information Stochastic Parity Games
Perfect-information Stochastic Parity Games Wies law Zielonka LIAFA, case 7014 Université Paris 7 2, Place Jussieu 75251 Paris Cedex 05, France zielonka@liafa.jussieu.fr Abstract. We show that in perfect-information
More informationThe Odds of Staying on Budget
The Odds of Staying on Budget Christoph Haase 1 and Stefan Kiefer 2 1 Laboratoire Spécification et Vérification (LSV), CNRS & ENS de Cachan, France 2 Department of Computer Science, University of Oxford,
More informationThe Complexity of Ergodic Mean-payoff Games,
The Complexity of Ergodic Mean-payoff Games, Krishnendu Chatterjee Rasmus Ibsen-Jensen Abstract We study two-player (zero-sum) concurrent mean-payoff games played on a finite-state graph. We focus on the
More informationNash Equilibria in Concurrent Games with Büchi Objectives
Nash Equilibria in Concurrent Games with Büchi Objectives Patricia Bouyer, Romain Brenguier, Nicolas Markey, and Michael Ummels LSV, CNRS & ENS Cachan, France {bouyer,brenguier,markey,ummels}@lsv.ens-cachan.fr
More informationGames with Discrete Resources
Games with Discrete Resources Sylvain Schmitz with Th. Colcombet, J.-B. Courtois, M. Jurdziński, and R. Lazić LSV, ENS Paris-Saclay & CNRS IBISC, October 19, 217 1/12 Outline multi-dimensional energy parity
More informationReachability-Time Games on Timed Automata (Extended Abstract)
Reachability-Time Games on Timed Automata (Extended Abstract) Marcin Jurdziński and Ashutosh Trivedi Department of Computer Science, University of Warwick, UK Abstract. In a reachability-time game, players
More informationLong-run Average Reward for Markov Decision Processes
Long-run Average Reward for Markov Decision Processes Based on a paper at CAV 2017 Pranav Ashok 1, Krishnendu Chatterjee 2, Przemysław Daca 2, Jan Křetínský 1 and Tobias Meggendorfer 1 August 9, 2017 1
More informationA lower bound for discounting algorithms solving two-person zero-sum limit average payoff stochastic games
R u t c o r Research R e p o r t A lower bound for discounting algorithms solving two-person zero-sum limit average payoff stochastic games Endre Boros a Vladimir Gurvich c Khaled Elbassioni b Kazuhisa
More informationProbabilistic verification and approximation schemes
Probabilistic verification and approximation schemes Richard Lassaigne Equipe de Logique mathématique, CNRS-Université Paris 7 Joint work with Sylvain Peyronnet (LRDE/EPITA & Equipe de Logique) Plan 1
More informationController Synthesis with Budget Constraints
Controller Synthesis with Budget Constraints Krishnendu Chatterjee 1, Rupak Majumdar 3, and Thomas A. Henzinger 1,2 1 EECS, UC Berkeley, 2 CCS, EPFL, 3 CS, UC Los Angeles Abstract. We study the controller
More informationSynthesis weakness of standard approach. Rational Synthesis
1 Synthesis weakness of standard approach Rational Synthesis 3 Overview Introduction to formal verification Reactive systems Verification Synthesis Introduction to Formal Verification of Reactive Systems
More informationInfinite-Duration Bidding Games
Infinite-Duration Bidding Games Guy Avni 1, Thomas A. Henzinger 2, and Ventsislav Chonev 3 1 IST Austria, Klosterneuburg, Austria 2 IST Austria, Klosterneuburg, Austria 3 Max Planck Institute for Software
More informationarxiv: v1 [cs.lo] 11 Jan 2019
Life is Random, Time is Not: Markov Decision Processes with Window Objectives Thomas Brihaye 1, Florent Delgrange 1,2, Youssouf Oualhadj 3, and Mickael Randour 1, 1 UMONS Université de Mons, Belgium 2
More informationPerfect-Information Stochastic Mean-Payo Parity Games
Perfect-Information Stochastic Mean-Payo Parity Games Krishnendu Chatterjee and Laurent Doyen and Hugo Gimbert and Youssouf Oualhadj Technical Report No. IST-2013-128-v1+1 Deposited at UNSPECIFIED http://repository.ist.ac.at/128/1/full
More informationA Gentle Introduction to Reinforcement Learning
A Gentle Introduction to Reinforcement Learning Alexander Jung 2018 1 Introduction and Motivation Consider the cleaning robot Rumba which has to clean the office room B329. In order to keep things simple,
More informationThe Stevens-Stirling-Algorithm For Solving Parity Games Locally Requires Exponential Time
The Stevens-Stirling-Algorithm For Solving Parity Games Locally Requires Exponential Time Oliver Friedmann Institut für Informatik, LMU München Oliver.Friedmann@googlemail.com Abstract. This paper presents
More informationOn the Power of Imperfect Information
Foundations of Software Technology and Theoretical Computer Science (Bangalore) 2008. Editors: R. Hariharan, M. Mukund, V. Vinay; pp 73-82 On the Power of Imperfect Information Dietmar Berwanger 1 and
More informationSynthesis from Probabilistic Components
Synthesis from Probabilistic Components Yoad Lustig, Sumit Nain, and Moshe Y. Vardi Department of Computer Science Rice University, Houston, TX 77005, USA yoad.lustig@gmail.com, nain@cs.rice.edu, vardi@cs.rice.edu
More informationQuantitative Languages
Quantitative Languages KRISHNENDU CHATTERJEE IST Austria (Institute of Science and Technology Austria) LAURENT DOYEN LSV, ENS Cachan & CNRS, France and THOMAS A. HENZINGER IST Austria (Institute of Science
More informationStrategy Synthesis for Markov Decision Processes and Branching-Time Logics
Strategy Synthesis for Markov Decision Processes and Branching-Time Logics Tomáš Brázdil and Vojtěch Forejt Faculty of Informatics, Masaryk University, Botanická 68a, 60200 Brno, Czech Republic. {brazdil,forejt}@fi.muni.cz
More informationProbabilistic Model Checking and Strategy Synthesis for Robot Navigation
Probabilistic Model Checking and Strategy Synthesis for Robot Navigation Dave Parker University of Birmingham (joint work with Bruno Lacerda, Nick Hawes) AIMS CDT, Oxford, May 2015 Overview Probabilistic
More informationChapter 4: Computation tree logic
INFOF412 Formal verification of computer systems Chapter 4: Computation tree logic Mickael Randour Formal Methods and Verification group Computer Science Department, ULB March 2017 1 CTL: a specification
More informationThe Complexity of Stochastic Müller Games
The Complexity of Stochastic Müller Games Krishnendu Chatterjee Electrical Engineering and Computer Sciences University of California at Berkeley Technical Report No. UCB/EECS-2007-110 http://www.eecs.berkeley.edu/pubs/techrpts/2007/eecs-2007-110.html
More informationLecture notes for Analysis of Algorithms : Markov decision processes
Lecture notes for Analysis of Algorithms : Markov decision processes Lecturer: Thomas Dueholm Hansen June 6, 013 Abstract We give an introduction to infinite-horizon Markov decision processes (MDPs) with
More informationRobust Reachability in Timed Automata: A Game-based Approach
Robust Reachability in Timed Automata: A Game-based Approach Patricia Bouyer, Nicolas Markey, and Ocan Sankur LSV, CNRS & ENS Cachan, France. {bouyer,markey,sankur}@lsv.ens-cachan.fr Abstract. Reachability
More informationFinitary Winning in ω-regular Games
Finitary Winning in ω-regular Games Krishnendu Chatterjee 1 and Thomas A. Henzinger 1,2 1 University of California, Berkeley, USA 2 EPFL, Switzerland {c krish,tah}@eecs.berkeley.edu Abstract. Games on
More informationConcept of Statistical Model Checking. Presented By: Diana EL RABIH
Concept of Statistical Model Checking Presented By: Diana EL RABIH Introduction Model checking of stochastic systems Continuous-time Markov chains Continuous Stochastic Logic (CSL) Probabilistic time-bounded
More informationValue Iteration. 1 Introduction. Krishnendu Chatterjee 1 and Thomas A. Henzinger 1,2
Value Iteration Krishnendu Chatterjee 1 and Thomas A. Henzinger 1,2 1 University of California, Berkeley 2 EPFL, Switzerland Abstract. We survey value iteration algorithms on graphs. Such algorithms can
More informationPerfect-Information Stochastic Mean-Payoff Parity Games,
Perfect-Information Stochastic Mean-Payoff Parity Games, Krishnendu Chatterjee (IST Austria), Laurent Doyen (LSV, ENS Cachan & CNRS, France), Hugo Gimbert (LaBRI, CNRS, France), and Youssouf Oualhadj (LIF,
More informationGames with Imperfect Information: Theory and Algorithms
Games with Imperfect Information: Theory and Algorithms Laurent Doyen 1 and Jean-François Raskin 2 1 LSV, ENS Cachan & CNRS, France 2 Université Libre de Bruxelles (ULB), Belgium Abstract. We study observation-based
More informationFinitary Winning in \omega-regular Games
Finitary Winning in \omega-regular Games Krishnendu Chatterjee Thomas A. Henzinger Florian Horn Electrical Engineering and Computer Sciences University of California at Berkeley Technical Report No. UCB/EECS-2007-120
More informationRobot Games of Degree Two
Robot Games of Degree Two V. Halava 1,2 R. Niskanen 2 I. Potapov 2 1 Department of Mathematics and Statistics University of Turku, Finland 2 Department of Computer Science University of Liverpool, UK 9th
More informationCentral-limit approach to risk-aware Markov decision processes
Central-limit approach to risk-aware Markov decision processes Jia Yuan Yu Concordia University November 27, 2015 Joint work with Pengqian Yu and Huan Xu. Inventory Management 1 1 Look at current inventory
More informationarxiv: v5 [cs.gt] 19 Feb 2018
Reactive Synthesis Without Regret Paul Hunter, Guillermo A. Pérez, and Jean-François Raskin Département d Informatique, Université Libre de Bruxelles {phunter,gperezme,jraskin}@ulb.ac.be February 2, 28
More informationStrategy Logic. 1 Introduction. Krishnendu Chatterjee 1, Thomas A. Henzinger 1,2, and Nir Piterman 2
Strategy Logic Krishnendu Chatterjee 1, Thomas A. Henzinger 1,2, and Nir Piterman 2 1 University of California, Berkeley, USA 2 EPFL, Switzerland c krish@eecs.berkeley.edu, {tah,nir.piterman}@epfl.ch Abstract.
More informationGeneralized Parity Games
Generalized Parity Games Krishnendu Chatterjee 1, Thomas A. Henzinger 1,2, and Nir Piterman 2 1 University of California, Berkeley, USA 2 EPFL, Switzerland c krish@eecs.berkeley.edu, {tah,nir.piterman}@epfl.ch
More informationTheory of Computation Chapter 1: Introduction
Theory of Computation Chapter 1: Introduction Guan-Shieng Huang Sep. 20, 2006 Feb. 9, 2009 0-0 Text Book Computational Complexity, by C. H. Papadimitriou, Addison-Wesley, 1994. 1 References Garey, M.R.
More informationA subexponential lower bound for the Random Facet algorithm for Parity Games
A subexponential lower bound for the Random Facet algorithm for Parity Games Oliver Friedmann 1 Thomas Dueholm Hansen 2 Uri Zwick 3 1 Department of Computer Science, University of Munich, Germany. 2 Center
More informationMDPs with Energy-Parity Objectives
MDPs with Energy-Parity Objectives Richard Mayr, Sven Schewe, Patrick Totzke, Dominik Wojtczak University of Edinburgh, UK University of Liverpool, UK Abstract Energy-parity objectives combine ω-regular
More informationQuantitative Languages
Quantitative Languages Krishnendu Chatterjee 1, Laurent Doyen 2, and Thomas A. Henzinger 2 1 University of California, Santa Cruz 2 EPFL, Lausanne, Switzerland Abstract. Quantitative generalizations of
More informationCS 781 Lecture 9 March 10, 2011 Topics: Local Search and Optimization Metropolis Algorithm Greedy Optimization Hopfield Networks Max Cut Problem Nash
CS 781 Lecture 9 March 10, 2011 Topics: Local Search and Optimization Metropolis Algorithm Greedy Optimization Hopfield Networks Max Cut Problem Nash Equilibrium Price of Stability Coping With NP-Hardness
More informationWhat s Decidable About Weighted Automata?
What s Decidable About Weighted Automata? Shaull Almagor 1, Udi Boker 1,2, and Orna Kupferman 1 1 Hebrew University, School of Engineering and Computer Science, Jerusalem, Israel. 2 IST, Austria. Abstract.
More informationNash Equilibria for Reachability Objectives in Multi-player Timed Games
Nash Equilibria for Reachability Objectives in Multi-player Timed Games Patricia Bouyer, Romain Brenguier, and Nicolas Markey LSV, ENS Cachan & CNRS, France {bouyer,brenguie,markey}@lsv.ens-cachan.fr Abstract.
More informationInformation Hiding and Covert Communication
Information Hiding and Covert Communication Andrew Ker adk @ comlab.ox.ac.uk Royal Society University Research Fellow Oxford University Computing Laboratory Foundations of Security Analysis and Design
More informationLimiting Behavior of Markov Chains with Eager Attractors
Limiting Behavior of Markov Chains with Eager Attractors Parosh Aziz Abdulla Uppsala University, Sweden. parosh@it.uu.se Noomene Ben Henda Uppsala University, Sweden. Noomene.BenHenda@it.uu.se Sven Sandberg
More information: Cryptography and Game Theory Ran Canetti and Alon Rosen. Lecture 8
0368.4170: Cryptography and Game Theory Ran Canetti and Alon Rosen Lecture 8 December 9, 2009 Scribe: Naama Ben-Aroya Last Week 2 player zero-sum games (min-max) Mixed NE (existence, complexity) ɛ-ne Correlated
More informationRobust Controller Synthesis in Timed Automata
Robust Controller Synthesis in Timed Automata Ocan Sankur LSV, ENS Cachan & CNRS Joint with Patricia Bouyer, Nicolas Markey, Pierre-Alain Reynier. Ocan Sankur (ENS Cachan) Robust Control in Timed Automata
More informationNP-Completeness. f(n) \ n n sec sec sec. n sec 24.3 sec 5.2 mins. 2 n sec 17.9 mins 35.
NP-Completeness Reference: Computers and Intractability: A Guide to the Theory of NP-Completeness by Garey and Johnson, W.H. Freeman and Company, 1979. NP-Completeness 1 General Problems, Input Size and
More informationSynthesis of Winning Strategies for Interaction under Partial Information
Synthesis of Winning Strategies for Interaction under Partial Information Oberseminar Informatik Bernd Puchala RWTH Aachen University June 10th, 2013 1 Introduction Interaction Strategy Synthesis 2 Main
More informationOptimism in the Face of Uncertainty Should be Refutable
Optimism in the Face of Uncertainty Should be Refutable Ronald ORTNER Montanuniversität Leoben Department Mathematik und Informationstechnolgie Franz-Josef-Strasse 18, 8700 Leoben, Austria, Phone number:
More informationOptimal Control of MDPs with Temporal Logic Constraints
52nd IEEE Conference on Decision and Control December 10-13, 2013. Florence, Italy Optimal Control of MDPs with Temporal Logic Constraints Mária Svoreňová, Ivana Černá and Calin Belta Abstract In this
More informationA Working Knowledge of Computational Complexity for an Optimizer
A Working Knowledge of Computational Complexity for an Optimizer ORF 363/COS 323 Instructor: Amir Ali Ahmadi 1 Why computational complexity? What is computational complexity theory? It s a branch of mathematics
More informationComputational Properties of Quasi-Strict Equilibrium
Computational roperties of Quasi-Strict Equilibrium Felix Brandt and Felix Fischer Institut für Informatik, Universität München 80538 München, Germany {brandtf,fischerf}@tcs.ifi.lmu.de Abstract. This paper
More informationPRISM An overview. automatic verification of systems with stochastic behaviour e.g. due to unreliability, uncertainty, randomisation,
PRISM An overview PRISM is a probabilistic model checker automatic verification of systems with stochastic behaviour e.g. due to unreliability, uncertainty, randomisation, Construction/analysis of probabilistic
More informationLearning to Coordinate Efficiently: A Model-based Approach
Journal of Artificial Intelligence Research 19 (2003) 11-23 Submitted 10/02; published 7/03 Learning to Coordinate Efficiently: A Model-based Approach Ronen I. Brafman Computer Science Department Ben-Gurion
More informationP versus NP. Math 40210, Spring September 16, Math (Spring 2012) P versus NP September 16, / 9
P versus NP Math 40210, Spring 2012 September 16, 2012 Math 40210 (Spring 2012) P versus NP September 16, 2012 1 / 9 Properties of graphs A property of a graph is anything that can be described without
More informationTwo Views on Multiple Mean-Payoff Objectives in Markov Decision Processes
wo Views on Multiple Mean-Payoff Objectives in Markov Decision Processes omáš Brázdil Faculty of Informatics Masaryk University Brno, Czech Republic brazdil@fi.muni.cz Václav Brožek LFCS, School of Informatics
More informationState Explosion in Almost-Sure Probabilistic Reachability
State Explosion in Almost-Sure Probabilistic Reachability François Laroussinie Lab. Spécification & Vérification, ENS de Cachan & CNRS UMR 8643, 61, av. Pdt. Wilson, 94235 Cachan Cedex France Jeremy Sproston
More informationEdinburgh Research Explorer
Edinburgh Research Explorer Recursive Concurrent Stochastic Games Citation for published version: Etessami, K & Yannakakis, M 2006, 'Recursive Concurrent Stochastic Games'. in ICALP (2). Springer- Verlag
More informationDuality in Robust Dynamic Programming: Pricing Convertibles, Stochastic Games and Control
Duality in Robust Dynamic Programming: Pricing Convertibles, Stochastic Games and Control Shyam S Chandramouli Abstract Many decision making problems that arise in inance, Economics, Inventory etc. can
More informationThe Complexity of Decentralized Control of Markov Decision Processes
The Complexity of Decentralized Control of Markov Decision Processes Daniel S. Bernstein Robert Givan Neil Immerman Shlomo Zilberstein Department of Computer Science University of Massachusetts Amherst,
More informationAntichain Algorithms for Finite Automata
Antichain Algorithms for Finite Automata Laurent Doyen 1 and Jean-François Raskin 2 1 LSV, ENS Cachan & CNRS, France 2 U.L.B., Université Libre de Bruxelles, Belgium Abstract. We present a general theory
More informationLecture December 2009 Fall 2009 Scribe: R. Ring In this lecture we will talk about
0368.4170: Cryptography and Game Theory Ran Canetti and Alon Rosen Lecture 7 02 December 2009 Fall 2009 Scribe: R. Ring In this lecture we will talk about Two-Player zero-sum games (min-max theorem) Mixed
More informationCS264: Beyond Worst-Case Analysis Lecture #18: Smoothed Complexity and Pseudopolynomial-Time Algorithms
CS264: Beyond Worst-Case Analysis Lecture #18: Smoothed Complexity and Pseudopolynomial-Time Algorithms Tim Roughgarden March 9, 2017 1 Preamble Our first lecture on smoothed analysis sought a better theoretical
More informationInfinite Games. Sumit Nain. 28 January Slides Credit: Barbara Jobstmann (CNRS/Verimag) Department of Computer Science Rice University
Infinite Games Sumit Nain Department of Computer Science Rice University 28 January 2013 Slides Credit: Barbara Jobstmann (CNRS/Verimag) Motivation Abstract games are of fundamental importance in mathematics
More informationTHE STEVENS-STIRLING-ALGORITHM FOR SOLVING PARITY GAMES LOCALLY REQUIRES EXPONENTIAL TIME
International Journal of Foundations of Computer Science c World Scientific Publishing Company THE STEVENS-STIRLING-ALGORITHM FOR SOLVING PARITY GAMES LOCALLY REQUIRES EXPONENTIAL TIME OLIVER FRIEDMANN
More informationReactive Synthesis. Swen Jacobs VTSA 2013 Nancy, France u
Reactive Synthesis Nancy, France 24.09.2013 u www.iaik.tugraz.at 2 Property Synthesis (You Will Never Code Again) 3 Construct Correct Systems Automatically Don t do the same
More informationQualitative Concurrent Parity Games
Qualitative Concurrent Parity Games Krishnendu Chatterjee Luca de Alfaro Thomas A Henzinger CE, University of California, Santa Cruz,USA EECS, University of California, Berkeley,USA Computer and Communication
More informationIntroduction to Game Theory
COMP323 Introduction to Computational Game Theory Introduction to Game Theory Paul G. Spirakis Department of Computer Science University of Liverpool Paul G. Spirakis (U. Liverpool) Introduction to Game
More information