approachability theory and differential games
play

Approachability Theory and Differential Games Sylvain Sorin - PowerPoint PPT Presentation

Approachability Theory and Differential Games Sylvain Sorin UPMC-Paris 6 Ecole Polytechnique sorin@math.jussieu.fr Sixi` emes Journ ees Franco-Chiliennes dOptimisation 19-21 mai 2008 Universit e du Sud Toulon-Var Sylvain Sorin


  1. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games H n = ( I × J ) n is the set of possible histories at stage n + 1 and H ∞ = ( I × J ) ∞ be the set of plays. Σ (resp. T ) is the set of strategies of Player 1 (resp. Player 2): mappings from H = ∪ n ≥ 0 H n to the sets of mixed actions U = ∆( I ) (probabilities on I ) (resp. V = ∆( J ) ). At stage n , given the history h n − 1 ∈ H n − 1 , Player 1 chooses an action in I according to the probability distribution σ ( h n − 1 ) ∈ U (and similarly for Player 2). A couple ( σ, τ ) of strategies induces a probability on H ∞ and E σ,τ denotes the corresponding expectation. Sylvain Sorin ATDG

  2. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games In the general theory of repeated two main approaches have been studied: - asympotic analysis: study of the limit of values of finitely repeated games or discounted games as the expected length goes to ∞ . This amounts to consider finer and finer time discretizations of a continuous time game played on [ 0 , 1 ] , - or uniform analysis through robustness properties of strategies : they should be approximately optimal in any sufficiently long game. In our framework this leads to the 2 following notions: Sylvain Sorin ATDG

  3. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games In the general theory of repeated two main approaches have been studied: - asympotic analysis: study of the limit of values of finitely repeated games or discounted games as the expected length goes to ∞ . This amounts to consider finer and finer time discretizations of a continuous time game played on [ 0 , 1 ] , - or uniform analysis through robustness properties of strategies : they should be approximately optimal in any sufficiently long game. In our framework this leads to the 2 following notions: Sylvain Sorin ATDG

  4. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games In the general theory of repeated two main approaches have been studied: - asympotic analysis: study of the limit of values of finitely repeated games or discounted games as the expected length goes to ∞ . This amounts to consider finer and finer time discretizations of a continuous time game played on [ 0 , 1 ] , - or uniform analysis through robustness properties of strategies : they should be approximately optimal in any sufficiently long game. In our framework this leads to the 2 following notions: Sylvain Sorin ATDG

  5. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games In the general theory of repeated two main approaches have been studied: - asympotic analysis: study of the limit of values of finitely repeated games or discounted games as the expected length goes to ∞ . This amounts to consider finer and finer time discretizations of a continuous time game played on [ 0 , 1 ] , - or uniform analysis through robustness properties of strategies : they should be approximately optimal in any sufficiently long game. In our framework this leads to the 2 following notions: Sylvain Sorin ATDG

  6. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games In the general theory of repeated two main approaches have been studied: - asympotic analysis: study of the limit of values of finitely repeated games or discounted games as the expected length goes to ∞ . This amounts to consider finer and finer time discretizations of a continuous time game played on [ 0 , 1 ] , - or uniform analysis through robustness properties of strategies : they should be approximately optimal in any sufficiently long game. In our framework this leads to the 2 following notions: Sylvain Sorin ATDG

  7. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games Definition R k is weakly approachable by A nonempty closed set C in I Player 1 in G if, for every ε > 0, there exists N ∈ I N such that for any n ≥ N there is a strategy σ = σ ( n , ε ) of Player 1 such that, for any strategy τ of Player 2 E σ,τ ( d C ( g n )) ≤ ε. where d C stands for the distance to C . If v n is the value of the n -stage game with payoff − E ( d C (¯ g n )) , weak-approachability means v n → 0. Sylvain Sorin ATDG

  8. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games Definition R k is weakly approachable by A nonempty closed set C in I Player 1 in G if, for every ε > 0, there exists N ∈ I N such that for any n ≥ N there is a strategy σ = σ ( n , ε ) of Player 1 such that, for any strategy τ of Player 2 E σ,τ ( d C ( g n )) ≤ ε. where d C stands for the distance to C . If v n is the value of the n -stage game with payoff − E ( d C (¯ g n )) , weak-approachability means v n → 0. Sylvain Sorin ATDG

  9. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games The uniform notion is Definition R k is approachable by Player 1 in A nonempty closed set C in I G if, for every ε > 0, there exists a strategy σ = σ ( ε ) of Player 1 and N ∈ I N such that, for any strategy τ of Player 2 and any n ≥ N E σ,τ ( d C ( g n )) ≤ ε. Asymptotically the average outcome remains close in expectation to the target C , uniformly with respect to the opponent’s behavior. Dual notion : excludability. Sylvain Sorin ATDG

  10. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games The uniform notion is Definition R k is approachable by Player 1 in A nonempty closed set C in I G if, for every ε > 0, there exists a strategy σ = σ ( ε ) of Player 1 and N ∈ I N such that, for any strategy τ of Player 2 and any n ≥ N E σ,τ ( d C ( g n )) ≤ ε. Asymptotically the average outcome remains close in expectation to the target C , uniformly with respect to the opponent’s behavior. Dual notion : excludability. Sylvain Sorin ATDG

  11. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games The model 1 Preliminaries 2 Weak approachability and differential games with fixed 3 duration Approachability and B -sets 4 Approachability and qualitative differential games 5 On strategies in the differential games and the repeated 6 games Sylvain Sorin ATDG

  12. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games 2.1 The “expected deterministic” repeated game form G ∗ Alternative two-person infinitely repeated game associated, as the previous one, to the matrix A . At each stage n = 1 , 2 , ... , Player 1 (resp. Player 2) chooses u n ∈ U = ∆( I ) (resp. v n ∈ V = ∆( J ) ), the outcome is g ∗ n = u n Av n and ( u n , v n ) is announced. Accordingly, a strategy σ ∗ for Player 1 in G ∗ is a map from H ∗ n = ( U × V ) n . A strategy τ ∗ for Player 2 = ∪ n ≥ 0 H ∗ n to U where H ∗ is defined similarly. A couple of strategies induces a play { ( u n , v n ) } and a sequence n } , and g ∗ n = 1 � n of outcomes { g ∗ m = 1 g ∗ m denotes the average n outcome up to stage n . G ∗ is the game played in “mixed moves” or in distribution. Weak ∗ approachability, v ∗ n and ∗ approachability are defined similarly. Sylvain Sorin ATDG

  13. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games 2.1 The “expected deterministic” repeated game form G ∗ Alternative two-person infinitely repeated game associated, as the previous one, to the matrix A . At each stage n = 1 , 2 , ... , Player 1 (resp. Player 2) chooses u n ∈ U = ∆( I ) (resp. v n ∈ V = ∆( J ) ), the outcome is g ∗ n = u n Av n and ( u n , v n ) is announced. Accordingly, a strategy σ ∗ for Player 1 in G ∗ is a map from H ∗ n = ( U × V ) n . A strategy τ ∗ for Player 2 = ∪ n ≥ 0 H ∗ n to U where H ∗ is defined similarly. A couple of strategies induces a play { ( u n , v n ) } and a sequence n } , and g ∗ n = 1 � n of outcomes { g ∗ m = 1 g ∗ m denotes the average n outcome up to stage n . G ∗ is the game played in “mixed moves” or in distribution. Weak ∗ approachability, v ∗ n and ∗ approachability are defined similarly. Sylvain Sorin ATDG

  14. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games 2.1 The “expected deterministic” repeated game form G ∗ Alternative two-person infinitely repeated game associated, as the previous one, to the matrix A . At each stage n = 1 , 2 , ... , Player 1 (resp. Player 2) chooses u n ∈ U = ∆( I ) (resp. v n ∈ V = ∆( J ) ), the outcome is g ∗ n = u n Av n and ( u n , v n ) is announced. Accordingly, a strategy σ ∗ for Player 1 in G ∗ is a map from H ∗ n = ( U × V ) n . A strategy τ ∗ for Player 2 = ∪ n ≥ 0 H ∗ n to U where H ∗ is defined similarly. A couple of strategies induces a play { ( u n , v n ) } and a sequence n } , and g ∗ n = 1 � n of outcomes { g ∗ m = 1 g ∗ m denotes the average n outcome up to stage n . G ∗ is the game played in “mixed moves” or in distribution. Weak ∗ approachability, v ∗ n and ∗ approachability are defined similarly. Sylvain Sorin ATDG

  15. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games 2.1 The “expected deterministic” repeated game form G ∗ Alternative two-person infinitely repeated game associated, as the previous one, to the matrix A . At each stage n = 1 , 2 , ... , Player 1 (resp. Player 2) chooses u n ∈ U = ∆( I ) (resp. v n ∈ V = ∆( J ) ), the outcome is g ∗ n = u n Av n and ( u n , v n ) is announced. Accordingly, a strategy σ ∗ for Player 1 in G ∗ is a map from H ∗ n = ( U × V ) n . A strategy τ ∗ for Player 2 = ∪ n ≥ 0 H ∗ n to U where H ∗ is defined similarly. A couple of strategies induces a play { ( u n , v n ) } and a sequence n } , and g ∗ n = 1 � n of outcomes { g ∗ m = 1 g ∗ m denotes the average n outcome up to stage n . G ∗ is the game played in “mixed moves” or in distribution. Weak ∗ approachability, v ∗ n and ∗ approachability are defined similarly. Sylvain Sorin ATDG

  16. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games 2.1 The “expected deterministic” repeated game form G ∗ Alternative two-person infinitely repeated game associated, as the previous one, to the matrix A . At each stage n = 1 , 2 , ... , Player 1 (resp. Player 2) chooses u n ∈ U = ∆( I ) (resp. v n ∈ V = ∆( J ) ), the outcome is g ∗ n = u n Av n and ( u n , v n ) is announced. Accordingly, a strategy σ ∗ for Player 1 in G ∗ is a map from H ∗ n = ( U × V ) n . A strategy τ ∗ for Player 2 = ∪ n ≥ 0 H ∗ n to U where H ∗ is defined similarly. A couple of strategies induces a play { ( u n , v n ) } and a sequence n } , and g ∗ n = 1 � n of outcomes { g ∗ m = 1 g ∗ m denotes the average n outcome up to stage n . G ∗ is the game played in “mixed moves” or in distribution. Weak ∗ approachability, v ∗ n and ∗ approachability are defined similarly. Sylvain Sorin ATDG

  17. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games 2.2 Differential games Consider zero-sum differential games of the kind ˙ x = f ( x , u , v ) where x is the state and u , v the moves. Assume R k ,  ( i ) U , V are compact subsets of I  R k × U × V �→ I R k  ( ii ) f : I is continuous,  ( iii ) f ( ., u , v ) is a l- Lipschitz map for all ( u , v ) ∈ U × V ,    ( iv ) U is convex, and f is affine in u . (1) Sylvain Sorin ATDG

  18. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games 2.2 Differential games Consider zero-sum differential games of the kind ˙ x = f ( x , u , v ) where x is the state and u , v the moves. Assume R k ,  ( i ) U , V are compact subsets of I  R k × U × V �→ I R k  ( ii ) f : I is continuous,  ( iii ) f ( ., u , v ) is a l- Lipschitz map for all ( u , v ) ∈ U × V ,    ( iv ) U is convex, and f is affine in u . (1) Sylvain Sorin ATDG

  19. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games 2.2 Differential games Consider zero-sum differential games of the kind ˙ x = f ( x , u , v ) where x is the state and u , v the moves. Assume R k ,  ( i ) U , V are compact subsets of I  R k × U × V �→ I R k  ( ii ) f : I is continuous,  ( iii ) f ( ., u , v ) is a l- Lipschitz map for all ( u , v ) ∈ U × V ,    ( iv ) U is convex, and f is affine in u . (1) Sylvain Sorin ATDG

  20. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games Sets of controls : U = { u : [ 0 , + ∞ ) �→ U ; u is measurable } and similarly V . R k and ( u , v ) ∈ U × V : Induced dynamics with x 0 ∈ I � ˙ x ( t ) = f ( x ( t ) , u ( t ) , v ( t )) for almost every t ≥ 0 (2) x ( 0 ) = x 0 . R k In addition Isaacs condition, namely : for any ζ ∈ I sup u ∈ U � ζ, f ( x , u , v ) � = inf inf u ∈ U sup � ζ, f ( x , u , v ) � . (3) v ∈ V v ∈ V Sylvain Sorin ATDG

  21. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games Sets of controls : U = { u : [ 0 , + ∞ ) �→ U ; u is measurable } and similarly V . R k and ( u , v ) ∈ U × V : Induced dynamics with x 0 ∈ I � ˙ x ( t ) = f ( x ( t ) , u ( t ) , v ( t )) for almost every t ≥ 0 (2) x ( 0 ) = x 0 . R k In addition Isaacs condition, namely : for any ζ ∈ I sup u ∈ U � ζ, f ( x , u , v ) � = inf inf u ∈ U sup � ζ, f ( x , u , v ) � . (3) v ∈ V v ∈ V Sylvain Sorin ATDG

  22. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games Sets of controls : U = { u : [ 0 , + ∞ ) �→ U ; u is measurable } and similarly V . R k and ( u , v ) ∈ U × V : Induced dynamics with x 0 ∈ I � ˙ x ( t ) = f ( x ( t ) , u ( t ) , v ( t )) for almost every t ≥ 0 (2) x ( 0 ) = x 0 . R k In addition Isaacs condition, namely : for any ζ ∈ I sup u ∈ U � ζ, f ( x , u , v ) � = inf inf u ∈ U sup � ζ, f ( x , u , v ) � . (3) v ∈ V v ∈ V Sylvain Sorin ATDG

  23. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games The model 1 Preliminaries 2 Weak approachability and differential games with fixed 3 duration Approachability and B -sets 4 Approachability and qualitative differential games 5 On strategies in the differential games and the repeated 6 games Sylvain Sorin ATDG

  24. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games (Vieille, 1992) The aim is to obtain a good average outcome at stage n . First consider the game G ∗ . Then use as state variable the cumulative payoff and consider the differential game Λ of fixed duration played on [ 0 , 1 ] starting from x ( 0 ) = 0 with dynamics: ˙ x ( t ) = u ( t ) A v ( t ) and payoff − d C ( x ( 1 )) . � t The state variable is x ( t ) = 0 g s ds with g s being the payoff at time s . G ∗ n appears then as a discrete time approximation of Λ . Let Φ( t , x ) be the value of the game played on [ t , 1 ] starting � 1 from x ( i.e. with total outcome x + t g s ds ). Sylvain Sorin ATDG

  25. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games (Vieille, 1992) The aim is to obtain a good average outcome at stage n . First consider the game G ∗ . Then use as state variable the cumulative payoff and consider the differential game Λ of fixed duration played on [ 0 , 1 ] starting from x ( 0 ) = 0 with dynamics: ˙ x ( t ) = u ( t ) A v ( t ) and payoff − d C ( x ( 1 )) . � t The state variable is x ( t ) = 0 g s ds with g s being the payoff at time s . G ∗ n appears then as a discrete time approximation of Λ . Let Φ( t , x ) be the value of the game played on [ t , 1 ] starting � 1 from x ( i.e. with total outcome x + t g s ds ). Sylvain Sorin ATDG

  26. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games (Vieille, 1992) The aim is to obtain a good average outcome at stage n . First consider the game G ∗ . Then use as state variable the cumulative payoff and consider the differential game Λ of fixed duration played on [ 0 , 1 ] starting from x ( 0 ) = 0 with dynamics: ˙ x ( t ) = u ( t ) A v ( t ) and payoff − d C ( x ( 1 )) . � t The state variable is x ( t ) = 0 g s ds with g s being the payoff at time s . G ∗ n appears then as a discrete time approximation of Λ . Let Φ( t , x ) be the value of the game played on [ t , 1 ] starting � 1 from x ( i.e. with total outcome x + t g s ds ). Sylvain Sorin ATDG

  27. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games (Vieille, 1992) The aim is to obtain a good average outcome at stage n . First consider the game G ∗ . Then use as state variable the cumulative payoff and consider the differential game Λ of fixed duration played on [ 0 , 1 ] starting from x ( 0 ) = 0 with dynamics: ˙ x ( t ) = u ( t ) A v ( t ) and payoff − d C ( x ( 1 )) . � t The state variable is x ( t ) = 0 g s ds with g s being the payoff at time s . G ∗ n appears then as a discrete time approximation of Λ . Let Φ( t , x ) be the value of the game played on [ t , 1 ] starting � 1 from x ( i.e. with total outcome x + t g s ds ). Sylvain Sorin ATDG

  28. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games (Vieille, 1992) The aim is to obtain a good average outcome at stage n . First consider the game G ∗ . Then use as state variable the cumulative payoff and consider the differential game Λ of fixed duration played on [ 0 , 1 ] starting from x ( 0 ) = 0 with dynamics: ˙ x ( t ) = u ( t ) A v ( t ) and payoff − d C ( x ( 1 )) . � t The state variable is x ( t ) = 0 g s ds with g s being the payoff at time s . G ∗ n appears then as a discrete time approximation of Λ . Let Φ( t , x ) be the value of the game played on [ t , 1 ] starting � 1 from x ( i.e. with total outcome x + t g s ds ). Sylvain Sorin ATDG

  29. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games (Vieille, 1992) The aim is to obtain a good average outcome at stage n . First consider the game G ∗ . Then use as state variable the cumulative payoff and consider the differential game Λ of fixed duration played on [ 0 , 1 ] starting from x ( 0 ) = 0 with dynamics: ˙ x ( t ) = u ( t ) A v ( t ) and payoff − d C ( x ( 1 )) . � t The state variable is x ( t ) = 0 g s ds with g s being the payoff at time s . G ∗ n appears then as a discrete time approximation of Λ . Let Φ( t , x ) be the value of the game played on [ t , 1 ] starting � 1 from x ( i.e. with total outcome x + t g s ds ). Sylvain Sorin ATDG

  30. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games Theorem 1) Φ( x , t ) is the unique viscosity solution of d dt Φ( x , t ) + val U × V �∇ Φ( x , t ) , uAv � = 0 with Φ( x , 1 ) = − d C ( x ) . 2) lim v ∗ n = Φ( 0 , 0 ) Sylvain Sorin ATDG

  31. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games The last step is to relate the values in G ∗ n and in G n . Theorem lim v ∗ n = lim v n Consider an optimal strategy in G ∗ n . Each stage m in this game will correspond to a block of L stages in G Ln where player 1 will play i.i.d. the prescribed strategy in G ∗ and will define inductively y ∗ m as the empirical distribution of moves of Player 2 during this block. Corollary Every set is weakly approachable or excludable. Sylvain Sorin ATDG

  32. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games The last step is to relate the values in G ∗ n and in G n . Theorem lim v ∗ n = lim v n Consider an optimal strategy in G ∗ n . Each stage m in this game will correspond to a block of L stages in G Ln where player 1 will play i.i.d. the prescribed strategy in G ∗ and will define inductively y ∗ m as the empirical distribution of moves of Player 2 during this block. Corollary Every set is weakly approachable or excludable. Sylvain Sorin ATDG

  33. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games The last step is to relate the values in G ∗ n and in G n . Theorem lim v ∗ n = lim v n Consider an optimal strategy in G ∗ n . Each stage m in this game will correspond to a block of L stages in G Ln where player 1 will play i.i.d. the prescribed strategy in G ∗ and will define inductively y ∗ m as the empirical distribution of moves of Player 2 during this block. Corollary Every set is weakly approachable or excludable. Sylvain Sorin ATDG

  34. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games The last step is to relate the values in G ∗ n and in G n . Theorem lim v ∗ n = lim v n Consider an optimal strategy in G ∗ n . Each stage m in this game will correspond to a block of L stages in G Ln where player 1 will play i.i.d. the prescribed strategy in G ∗ and will define inductively y ∗ m as the empirical distribution of moves of Player 2 during this block. Corollary Every set is weakly approachable or excludable. Sylvain Sorin ATDG

  35. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games The last step is to relate the values in G ∗ n and in G n . Theorem lim v ∗ n = lim v n Consider an optimal strategy in G ∗ n . Each stage m in this game will correspond to a block of L stages in G Ln where player 1 will play i.i.d. the prescribed strategy in G ∗ and will define inductively y ∗ m as the empirical distribution of moves of Player 2 during this block. Corollary Every set is weakly approachable or excludable. Sylvain Sorin ATDG

  36. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games The model 1 Preliminaries 2 Weak approachability and differential games with fixed 3 duration Approachability and B -sets 4 Approachability and qualitative differential games 5 On strategies in the differential games and the repeated 6 games Sylvain Sorin ATDG

  37. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games The main notion was introduced by Blackwell: Definition R k is a B -set for Player 1 (given A ), if for any A closed set C in I ∈ C , there exists y ∈ π C ( z ) and a mixed action u = ˆ z / u ( z ) in U = ∆( I ) such that the hyperplane through y orthogonal to the segment [ yz ] separates z from uAV : � uAv − y , z − y � ≤ 0 , ∀ v ∈ V . where π C ( z ) denotes the set of closest points to z in C . Sylvain Sorin ATDG

  38. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games The main notion was introduced by Blackwell: Definition R k is a B -set for Player 1 (given A ), if for any A closed set C in I ∈ C , there exists y ∈ π C ( z ) and a mixed action u = ˆ z / u ( z ) in U = ∆( I ) such that the hyperplane through y orthogonal to the segment [ yz ] separates z from uAV : � uAv − y , z − y � ≤ 0 , ∀ v ∈ V . where π C ( z ) denotes the set of closest points to z in C . Sylvain Sorin ATDG

  39. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games The basic Blackwell’s result is : Theorem Let C be a B -set for Player 1. Then it is approachable in G and ∗ approachable in G ∗ by that player. An approchability strategy is given by σ ( h n ) = ˆ u (¯ g n ) (resp. σ ∗ ( h ∗ n ) = ˆ u (¯ g ∗ n )) . The proof for approachability is Proposition 8 in [1]. The other one is a simple adaptation where the outcome ¯ g n is replaced by ¯ g ∗ n . Remark. The previous Proposition implies that a B -set remains approachable (resp. ∗ approachable) in the game where the only information of Player 1 after stage n is the current outcome g n (resp. g ∗ n ) rather than the complete previous history h n (resp. h ∗ n ). (natural state variable) Sylvain Sorin ATDG

  40. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games The basic Blackwell’s result is : Theorem Let C be a B -set for Player 1. Then it is approachable in G and ∗ approachable in G ∗ by that player. An approchability strategy is given by σ ( h n ) = ˆ u (¯ g n ) (resp. σ ∗ ( h ∗ n ) = ˆ u (¯ g ∗ n )) . The proof for approachability is Proposition 8 in [1]. The other one is a simple adaptation where the outcome ¯ g n is replaced by ¯ g ∗ n . Remark. The previous Proposition implies that a B -set remains approachable (resp. ∗ approachable) in the game where the only information of Player 1 after stage n is the current outcome g n (resp. g ∗ n ) rather than the complete previous history h n (resp. h ∗ n ). (natural state variable) Sylvain Sorin ATDG

  41. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games The basic Blackwell’s result is : Theorem Let C be a B -set for Player 1. Then it is approachable in G and ∗ approachable in G ∗ by that player. An approchability strategy is given by σ ( h n ) = ˆ u (¯ g n ) (resp. σ ∗ ( h ∗ n ) = ˆ u (¯ g ∗ n )) . The proof for approachability is Proposition 8 in [1]. The other one is a simple adaptation where the outcome ¯ g n is replaced by ¯ g ∗ n . Remark. The previous Proposition implies that a B -set remains approachable (resp. ∗ approachable) in the game where the only information of Player 1 after stage n is the current outcome g n (resp. g ∗ n ) rather than the complete previous history h n (resp. h ∗ n ). (natural state variable) Sylvain Sorin ATDG

  42. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games An important consequence of this property is Theorem A convex set C is either approachable or excludable. A further result due to Spinat [11] characterizes minimal approachable sets: Theorem A set C is approachable iff it contains a B -set. Sylvain Sorin ATDG

  43. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games The model 1 Preliminaries 2 Weak approachability and differential games with fixed 3 duration Approachability and B -sets 4 Approachability and qualitative differential games 5 On strategies in the differential games and the repeated 6 games Sylvain Sorin ATDG

  44. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games To study ∗ approachability, we introduce an allternative differential game Γ where both the dynamics and the payoff differs from the previous differential game Λ . The aim is to control the average payoff hence the discrete dynamics on the state variable is of the form 1 g n + 1 − ¯ ¯ n + 1 ( g n + 1 − ¯ g n = g n ) � t Continuous counterpart is γ ( u , v )( t ) = 1 0 u ( s ) A v ( s ) ds . t Change of variable x ( s ) = γ ( e s ) leads to ˙ x ( t ) = u ( t ) A v ( t ) − x ( t ) . (4) which is the dynamics of a differential game Γ with f ( x , u , v ) = uAv − x Sylvain Sorin ATDG

  45. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games To study ∗ approachability, we introduce an allternative differential game Γ where both the dynamics and the payoff differs from the previous differential game Λ . The aim is to control the average payoff hence the discrete dynamics on the state variable is of the form 1 g n + 1 − ¯ ¯ n + 1 ( g n + 1 − ¯ g n = g n ) � t Continuous counterpart is γ ( u , v )( t ) = 1 0 u ( s ) A v ( s ) ds . t Change of variable x ( s ) = γ ( e s ) leads to ˙ x ( t ) = u ( t ) A v ( t ) − x ( t ) . (4) which is the dynamics of a differential game Γ with f ( x , u , v ) = uAv − x Sylvain Sorin ATDG

  46. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games To study ∗ approachability, we introduce an allternative differential game Γ where both the dynamics and the payoff differs from the previous differential game Λ . The aim is to control the average payoff hence the discrete dynamics on the state variable is of the form 1 g n + 1 − ¯ ¯ n + 1 ( g n + 1 − ¯ g n = g n ) � t Continuous counterpart is γ ( u , v )( t ) = 1 0 u ( s ) A v ( s ) ds . t Change of variable x ( s ) = γ ( e s ) leads to ˙ x ( t ) = u ( t ) A v ( t ) − x ( t ) . (4) which is the dynamics of a differential game Γ with f ( x , u , v ) = uAv − x Sylvain Sorin ATDG

  47. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games To study ∗ approachability, we introduce an allternative differential game Γ where both the dynamics and the payoff differs from the previous differential game Λ . The aim is to control the average payoff hence the discrete dynamics on the state variable is of the form 1 g n + 1 − ¯ ¯ n + 1 ( g n + 1 − ¯ g n = g n ) � t Continuous counterpart is γ ( u , v )( t ) = 1 0 u ( s ) A v ( s ) ds . t Change of variable x ( s ) = γ ( e s ) leads to ˙ x ( t ) = u ( t ) A v ( t ) − x ( t ) . (4) which is the dynamics of a differential game Γ with f ( x , u , v ) = uAv − x Sylvain Sorin ATDG

  48. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games In addition the aim of Player 1 is to stay in a certain set C . We introduce the following definitions: Definition A map α : V → U is a nonanticipative strategy if, for any t ≥ 0 and for any v 1 and v 2 of V , which coincide almost everywhere on [ 0 , t ] of [ 0 , + ∞ ) , the images α ( v 1 ) and α ( v 2 ) coincide also almost everywhere on [ 0 , t ] . M ( V , U ) is the set of nonanticipative strategies from V to U . Sylvain Sorin ATDG

  49. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games In addition the aim of Player 1 is to stay in a certain set C . We introduce the following definitions: Definition A map α : V → U is a nonanticipative strategy if, for any t ≥ 0 and for any v 1 and v 2 of V , which coincide almost everywhere on [ 0 , t ] of [ 0 , + ∞ ) , the images α ( v 1 ) and α ( v 2 ) coincide also almost everywhere on [ 0 , t ] . M ( V , U ) is the set of nonanticipative strategies from V to U . Sylvain Sorin ATDG

  50. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games In addition the aim of Player 1 is to stay in a certain set C . We introduce the following definitions: Definition A map α : V → U is a nonanticipative strategy if, for any t ≥ 0 and for any v 1 and v 2 of V , which coincide almost everywhere on [ 0 , t ] of [ 0 , + ∞ ) , the images α ( v 1 ) and α ( v 2 ) coincide also almost everywhere on [ 0 , t ] . M ( V , U ) is the set of nonanticipative strategies from V to U . Sylvain Sorin ATDG

  51. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games Definition R k is a discriminating domain A non-empty closed set C in I for Player 1, given f if: ∀ x ∈ C , ∀ p ∈ NP C ( x ) , sup u ∈ U � f ( x , u , v ) , p � ≤ 0 , inf (5) v ∈ V where NP C ( x ) is the set of proximal normals to C at x R K ; d C ( x + p ) = � p �} NP C ( x ) = { p ∈ I The interpretation is that, at any boundary point x ∈ C , Player 1 can react to any control of Player 2 in order to keep the trajectory in the half space facing a proximal normal p . Sylvain Sorin ATDG

  52. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games Definition R k is a discriminating domain A non-empty closed set C in I for Player 1, given f if: ∀ x ∈ C , ∀ p ∈ NP C ( x ) , sup u ∈ U � f ( x , u , v ) , p � ≤ 0 , inf (5) v ∈ V where NP C ( x ) is the set of proximal normals to C at x R K ; d C ( x + p ) = � p �} NP C ( x ) = { p ∈ I The interpretation is that, at any boundary point x ∈ C , Player 1 can react to any control of Player 2 in order to keep the trajectory in the half space facing a proximal normal p . Sylvain Sorin ATDG

  53. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games Definition R k is a discriminating domain A non-empty closed set C in I for Player 1, given f if: ∀ x ∈ C , ∀ p ∈ NP C ( x ) , sup u ∈ U � f ( x , u , v ) , p � ≤ 0 , inf (5) v ∈ V where NP C ( x ) is the set of proximal normals to C at x R K ; d C ( x + p ) = � p �} NP C ( x ) = { p ∈ I The interpretation is that, at any boundary point x ∈ C , Player 1 can react to any control of Player 2 in order to keep the trajectory in the half space facing a proximal normal p . Sylvain Sorin ATDG

  54. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games The following theorem, due to Cardaliaguet [2], states that Player 1 can ensure remaining in a discriminating domain as soon as he knows, at each time t , Player 2’s control up to time t . Theorem Assume that f satisfies conditions (1), and that C is a closed R k . Then C is a discriminating domain if and only if subset of I for every x 0 belonging to C, there exists a nonanticipative strategy α ∈ M ( V , U ) , such that for any v ∈ V , the solution x [ x 0 , α ( v ) , v ]( t ) remains in C for every t ≥ 0 . We shall say that such a strategy α preserves the set C . Sylvain Sorin ATDG

  55. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games The following theorem, due to Cardaliaguet [2], states that Player 1 can ensure remaining in a discriminating domain as soon as he knows, at each time t , Player 2’s control up to time t . Theorem Assume that f satisfies conditions (1), and that C is a closed R k . Then C is a discriminating domain if and only if subset of I for every x 0 belonging to C, there exists a nonanticipative strategy α ∈ M ( V , U ) , such that for any v ∈ V , the solution x [ x 0 , α ( v ) , v ]( t ) remains in C for every t ≥ 0 . We shall say that such a strategy α preserves the set C . Sylvain Sorin ATDG

  56. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games Theorem R k is a Let f ( x , u , v ) = uAv − x. A closed set C ⊂ I discriminating domain for Player 1, if and only if C is a B -set for Player 1. First condition: start from z , x = π C ( z ) , there exists u such that for all v � uAv − x , z − x � ≤ 0. Second condition: start from x and p ∈ NP C ( x ) , then sup u ∈ U � f ( x , u , v ) , p � ≤ 0 . inf v ∈ V Sylvain Sorin ATDG

  57. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games Theorem R k is a Let f ( x , u , v ) = uAv − x. A closed set C ⊂ I discriminating domain for Player 1, if and only if C is a B -set for Player 1. First condition: start from z , x = π C ( z ) , there exists u such that for all v � uAv − x , z − x � ≤ 0. Second condition: start from x and p ∈ NP C ( x ) , then sup u ∈ U � f ( x , u , v ) , p � ≤ 0 . inf v ∈ V Sylvain Sorin ATDG

  58. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games Theorem R k is a Let f ( x , u , v ) = uAv − x. A closed set C ⊂ I discriminating domain for Player 1, if and only if C is a B -set for Player 1. First condition: start from z , x = π C ( z ) , there exists u such that for all v � uAv − x , z − x � ≤ 0. Second condition: start from x and p ∈ NP C ( x ) , then sup u ∈ U � f ( x , u , v ) , p � ≤ 0 . inf v ∈ V Sylvain Sorin ATDG

  59. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games It is easy to deduce that starting from any point, not necessarily in C one has: Theorem R k is a B -set for Player 1, there exists a If a closed set C ⊂ I nonanticipative strategy of player 1 in Γ , α ∈ M ( V , U ) , such that for every v ∈ V d C ( x [ α ( v ) , v ]( t )) ≤ Me − t . ∀ t ≥ 1 (6) Sylvain Sorin ATDG

  60. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games Theorem A closed set C is ∗ approachable for Player 1 in G ∗ if and only if it contains a B -set for Player 1 (given A). The direct part follows from Blackwell’s proof. To obtain the converse implication, the proof follows several steps: First, we construct a map Ψ from strategies of Player 1 in G ∗ to nonanticipative strategies in Γ . In particular given ε > 0 and a strategy σ ε that ε -approaches C in G ∗ , we define its image α ε = Ψ( σ ε ) . The next step consists in proving that the trajectories in the differential game Γ compatible with α ε approach asymptotically C + ε ¯ B . Sylvain Sorin ATDG

  61. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games Theorem A closed set C is ∗ approachable for Player 1 in G ∗ if and only if it contains a B -set for Player 1 (given A). The direct part follows from Blackwell’s proof. To obtain the converse implication, the proof follows several steps: First, we construct a map Ψ from strategies of Player 1 in G ∗ to nonanticipative strategies in Γ . In particular given ε > 0 and a strategy σ ε that ε -approaches C in G ∗ , we define its image α ε = Ψ( σ ε ) . The next step consists in proving that the trajectories in the differential game Γ compatible with α ε approach asymptotically C + ε ¯ B . Sylvain Sorin ATDG

  62. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games Theorem A closed set C is ∗ approachable for Player 1 in G ∗ if and only if it contains a B -set for Player 1 (given A). The direct part follows from Blackwell’s proof. To obtain the converse implication, the proof follows several steps: First, we construct a map Ψ from strategies of Player 1 in G ∗ to nonanticipative strategies in Γ . In particular given ε > 0 and a strategy σ ε that ε -approaches C in G ∗ , we define its image α ε = Ψ( σ ε ) . The next step consists in proving that the trajectories in the differential game Γ compatible with α ε approach asymptotically C + ε ¯ B . Sylvain Sorin ATDG

  63. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games Theorem A closed set C is ∗ approachable for Player 1 in G ∗ if and only if it contains a B -set for Player 1 (given A). The direct part follows from Blackwell’s proof. To obtain the converse implication, the proof follows several steps: First, we construct a map Ψ from strategies of Player 1 in G ∗ to nonanticipative strategies in Γ . In particular given ε > 0 and a strategy σ ε that ε -approaches C in G ∗ , we define its image α ε = Ψ( σ ε ) . The next step consists in proving that the trajectories in the differential game Γ compatible with α ε approach asymptotically C + ε ¯ B . Sylvain Sorin ATDG

  64. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games Theorem A closed set C is ∗ approachable for Player 1 in G ∗ if and only if it contains a B -set for Player 1 (given A). The direct part follows from Blackwell’s proof. To obtain the converse implication, the proof follows several steps: First, we construct a map Ψ from strategies of Player 1 in G ∗ to nonanticipative strategies in Γ . In particular given ε > 0 and a strategy σ ε that ε -approaches C in G ∗ , we define its image α ε = Ψ( σ ε ) . The next step consists in proving that the trajectories in the differential game Γ compatible with α ε approach asymptotically C + ε ¯ B . Sylvain Sorin ATDG

  65. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games Theorem A closed set C is ∗ approachable for Player 1 in G ∗ if and only if it contains a B -set for Player 1 (given A). The direct part follows from Blackwell’s proof. To obtain the converse implication, the proof follows several steps: First, we construct a map Ψ from strategies of Player 1 in G ∗ to nonanticipative strategies in Γ . In particular given ε > 0 and a strategy σ ε that ε -approaches C in G ∗ , we define its image α ε = Ψ( σ ε ) . The next step consists in proving that the trajectories in the differential game Γ compatible with α ε approach asymptotically C + ε ¯ B . Sylvain Sorin ATDG

  66. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games Then, we show that the ω -limit set of any trajectory compatible with some α ∈ M ( V , U ) is a nonempty compact discriminating domain for f . Explicitely, let � D ( α ) = cl { x [ x 0 , α ( w ) , w ]( t ); t ≥ θ, w ∈ V } . θ ≥ 0 (where cl is the closure operator). Lemma D ( α ) is a nonempty compact discriminating domain for Player 1 given f. Sylvain Sorin ATDG

  67. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games Then, we show that the ω -limit set of any trajectory compatible with some α ∈ M ( V , U ) is a nonempty compact discriminating domain for f . Explicitely, let � D ( α ) = cl { x [ x 0 , α ( w ) , w ]( t ); t ≥ θ, w ∈ V } . θ ≥ 0 (where cl is the closure operator). Lemma D ( α ) is a nonempty compact discriminating domain for Player 1 given f. Sylvain Sorin ATDG

  68. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games Then, we show that the ω -limit set of any trajectory compatible with some α ∈ M ( V , U ) is a nonempty compact discriminating domain for f . Explicitely, let � D ( α ) = cl { x [ x 0 , α ( w ) , w ]( t ); t ≥ θ, w ∈ V } . θ ≥ 0 (where cl is the closure operator). Lemma D ( α ) is a nonempty compact discriminating domain for Player 1 given f. Sylvain Sorin ATDG

  69. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games Recall that Spinat [11] proved that a closed set C is approachable in G if and only if it contains a B -set, hence we deduce the following corollary. Corollary Approachability and ∗ approachability coincide. Sylvain Sorin ATDG

  70. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games The model 1 Preliminaries 2 Weak approachability and differential games with fixed 3 duration Approachability and B -sets 4 Approachability and qualitative differential games 5 On strategies in the differential games and the repeated 6 games Sylvain Sorin ATDG

  71. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games First introduce a new notion of strategies crucial for time discretization. Definition A map δ : V �→ U is a nonanticipative strategy with delay (NAD) if there exits a sequence of times 0 < t 1 < t 2 < .. < t n < .. going to ∞ with the following property : For every control v 1 , v 2 ∈ U such that v 1 ( s ) = v 2 ( s ) for almost every s ∈ [ 0 , t i ] then δ ( v 1 )( s ) = δ ( v 2 )( s ) for almost every s ∈ [ 0 , t i + 1 ] . Denote by M d ( V , U ) the set of such nonanticipative strategies with delay from V to U . Sylvain Sorin ATDG

  72. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games First introduce a new notion of strategies crucial for time discretization. Definition A map δ : V �→ U is a nonanticipative strategy with delay (NAD) if there exits a sequence of times 0 < t 1 < t 2 < .. < t n < .. going to ∞ with the following property : For every control v 1 , v 2 ∈ U such that v 1 ( s ) = v 2 ( s ) for almost every s ∈ [ 0 , t i ] then δ ( v 1 )( s ) = δ ( v 2 )( s ) for almost every s ∈ [ 0 , t i + 1 ] . Denote by M d ( V , U ) the set of such nonanticipative strategies with delay from V to U . Sylvain Sorin ATDG

  73. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games First introduce a new notion of strategies crucial for time discretization. Definition A map δ : V �→ U is a nonanticipative strategy with delay (NAD) if there exits a sequence of times 0 < t 1 < t 2 < .. < t n < .. going to ∞ with the following property : For every control v 1 , v 2 ∈ U such that v 1 ( s ) = v 2 ( s ) for almost every s ∈ [ 0 , t i ] then δ ( v 1 )( s ) = δ ( v 2 )( s ) for almost every s ∈ [ 0 , t i + 1 ] . Denote by M d ( V , U ) the set of such nonanticipative strategies with delay from V to U . Sylvain Sorin ATDG

  74. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games We establish a link between preserving NA strategies in the differential game Γ and approachability strategies in the repeated game G . The idea of the construction is the following: a) Given a NA strategy α , show that it can be approximated in term of range by a NAD strategy ¯ α . b) When applied to α preserving C (hence approaching C ), obtain a NAD strategy ¯ α approaching C . c) This NAD strategy ¯ α produces an ∗ approachability strategy in the repeated game G ∗ . d) Finally ∗ approachability strategies in G ∗ induce approachability strategies in G . Sylvain Sorin ATDG

  75. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games We establish a link between preserving NA strategies in the differential game Γ and approachability strategies in the repeated game G . The idea of the construction is the following: a) Given a NA strategy α , show that it can be approximated in term of range by a NAD strategy ¯ α . b) When applied to α preserving C (hence approaching C ), obtain a NAD strategy ¯ α approaching C . c) This NAD strategy ¯ α produces an ∗ approachability strategy in the repeated game G ∗ . d) Finally ∗ approachability strategies in G ∗ induce approachability strategies in G . Sylvain Sorin ATDG

  76. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games We establish a link between preserving NA strategies in the differential game Γ and approachability strategies in the repeated game G . The idea of the construction is the following: a) Given a NA strategy α , show that it can be approximated in term of range by a NAD strategy ¯ α . b) When applied to α preserving C (hence approaching C ), obtain a NAD strategy ¯ α approaching C . c) This NAD strategy ¯ α produces an ∗ approachability strategy in the repeated game G ∗ . d) Finally ∗ approachability strategies in G ∗ induce approachability strategies in G . Sylvain Sorin ATDG

  77. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games We establish a link between preserving NA strategies in the differential game Γ and approachability strategies in the repeated game G . The idea of the construction is the following: a) Given a NA strategy α , show that it can be approximated in term of range by a NAD strategy ¯ α . b) When applied to α preserving C (hence approaching C ), obtain a NAD strategy ¯ α approaching C . c) This NAD strategy ¯ α produces an ∗ approachability strategy in the repeated game G ∗ . d) Finally ∗ approachability strategies in G ∗ induce approachability strategies in G . Sylvain Sorin ATDG

  78. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games We establish a link between preserving NA strategies in the differential game Γ and approachability strategies in the repeated game G . The idea of the construction is the following: a) Given a NA strategy α , show that it can be approximated in term of range by a NAD strategy ¯ α . b) When applied to α preserving C (hence approaching C ), obtain a NAD strategy ¯ α approaching C . c) This NAD strategy ¯ α produces an ∗ approachability strategy in the repeated game G ∗ . d) Finally ∗ approachability strategies in G ∗ induce approachability strategies in G . Sylvain Sorin ATDG

  79. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games Step a. Range associated to a nonanticipative strategy α ∈ M ( V , U ) : R k ∃ v ∈ V , y = x [ x 0 , α ( v ) , v ]( t ) } . R ( α, t ) = cl { y ∈ I The next result is due to Cardaliaguet ([4]) and is inspired by the ”extremal aiming” method of Krasowkii and Subbotin [9], and is very much in the spirit of proximal normals and approachability. Sylvain Sorin ATDG

  80. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games Step a. Range associated to a nonanticipative strategy α ∈ M ( V , U ) : R k ∃ v ∈ V , y = x [ x 0 , α ( v ) , v ]( t ) } . R ( α, t ) = cl { y ∈ I The next result is due to Cardaliaguet ([4]) and is inspired by the ”extremal aiming” method of Krasowkii and Subbotin [9], and is very much in the spirit of proximal normals and approachability. Sylvain Sorin ATDG

  81. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games Proposition Consider the differential game (2). For any ε > 0 , T > 0 and any nonanticipative strategy α ∈ M ( V , U ) , there exists some nonanticipative strategy with delay α ∈ M d ( V , U ) such that, for all t ∈ [ 0 , T ] and all v ∈ V : d R ( α, t ) ( x [ x 0 , α ( v ) , v ]( t )) ≤ ε. Assume that x k does not belong to R ( α, t k ) . Then there exists some control v k ∈ V such that y k := x [ t 0 , x 0 , α ( v k ) , v k ]( t k ) is an approximate closest point to x k in R ( α, t k ) . Note p k := x k − y k and take u k ∈ U such that < f ( x k , u k , v ) , p k > = inf < f ( x k , u , v ) , p k > = A k . (7) sup u ∈ U sup v ∈ V v ∈ V In words, u k is optimal in the local game at x k in direction p k . Sylvain Sorin ATDG

  82. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games Proposition Consider the differential game (2). For any ε > 0 , T > 0 and any nonanticipative strategy α ∈ M ( V , U ) , there exists some nonanticipative strategy with delay α ∈ M d ( V , U ) such that, for all t ∈ [ 0 , T ] and all v ∈ V : d R ( α, t ) ( x [ x 0 , α ( v ) , v ]( t )) ≤ ε. Assume that x k does not belong to R ( α, t k ) . Then there exists some control v k ∈ V such that y k := x [ t 0 , x 0 , α ( v k ) , v k ]( t k ) is an approximate closest point to x k in R ( α, t k ) . Note p k := x k − y k and take u k ∈ U such that < f ( x k , u k , v ) , p k > = inf < f ( x k , u , v ) , p k > = A k . (7) sup u ∈ U sup v ∈ V v ∈ V In words, u k is optimal in the local game at x k in direction p k . Sylvain Sorin ATDG

  83. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games Proposition Consider the differential game (2). For any ε > 0 , T > 0 and any nonanticipative strategy α ∈ M ( V , U ) , there exists some nonanticipative strategy with delay α ∈ M d ( V , U ) such that, for all t ∈ [ 0 , T ] and all v ∈ V : d R ( α, t ) ( x [ x 0 , α ( v ) , v ]( t )) ≤ ε. Assume that x k does not belong to R ( α, t k ) . Then there exists some control v k ∈ V such that y k := x [ t 0 , x 0 , α ( v k ) , v k ]( t k ) is an approximate closest point to x k in R ( α, t k ) . Note p k := x k − y k and take u k ∈ U such that < f ( x k , u k , v ) , p k > = inf < f ( x k , u , v ) , p k > = A k . (7) sup u ∈ U sup v ∈ V v ∈ V In words, u k is optimal in the local game at x k in direction p k . Sylvain Sorin ATDG

  84. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games Proposition Consider the differential game (2). For any ε > 0 , T > 0 and any nonanticipative strategy α ∈ M ( V , U ) , there exists some nonanticipative strategy with delay α ∈ M d ( V , U ) such that, for all t ∈ [ 0 , T ] and all v ∈ V : d R ( α, t ) ( x [ x 0 , α ( v ) , v ]( t )) ≤ ε. Assume that x k does not belong to R ( α, t k ) . Then there exists some control v k ∈ V such that y k := x [ t 0 , x 0 , α ( v k ) , v k ]( t k ) is an approximate closest point to x k in R ( α, t k ) . Note p k := x k − y k and take u k ∈ U such that < f ( x k , u k , v ) , p k > = inf < f ( x k , u , v ) , p k > = A k . (7) sup u ∈ U sup v ∈ V v ∈ V In words, u k is optimal in the local game at x k in direction p k . Sylvain Sorin ATDG

  85. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games The next result relies explicitely on the specific form (4) of the dynamics f in Γ and extends the approximation from a compact R + . interval to to I Proposition R k . For any ε > 0 and any nonanticipative strategy Fix x 0 ∈ I α ∈ M ( V , U ) in the game Γ , there is some nonanticipative strategy with delay α ∈ M d ( V , U ) such that, for all t ≥ 0 and all v ∈ V : d R ( α, t ) ( x [ x 0 , α ( v ) , v ]( t )) ≤ ε. Sylvain Sorin ATDG

  86. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games The next result relies explicitely on the specific form (4) of the dynamics f in Γ and extends the approximation from a compact R + . interval to to I Proposition R k . For any ε > 0 and any nonanticipative strategy Fix x 0 ∈ I α ∈ M ( V , U ) in the game Γ , there is some nonanticipative strategy with delay α ∈ M d ( V , U ) such that, for all t ≥ 0 and all v ∈ V : d R ( α, t ) ( x [ x 0 , α ( v ) , v ]( t )) ≤ ε. Sylvain Sorin ATDG

  87. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games In particular, step b) Proposition Let C be a B -set. For any ε > 0 there is some nonanticipative strategy with delay α ∈ M d ( V , U ) in the game Γ and some T such that for any v in V d C ( γ [ α ( v ) , v ]( t )) ≤ ε, ∀ t ≥ T . Sylvain Sorin ATDG

  88. The model Preliminaries Weak approachability and differential games with fixed duration Approachability and B-sets Approachability and qualitative differential games On strategies in the differential games and the repeated games step c) Proposition For any ε > 0 and any nonanticipative strategy α ∈ M ( V , U ) preserving C in the game Γ , there is some nonanticipative strategy with delay α ∈ M d ( V , U ) that induces an ε -approachability strategy σ ∗ for C in G ∗ . Idea is to use the delay to define a strategy that depens only on the past moves. Sylvain Sorin ATDG

Recommend


More recommend