an introduction to cooperative game theory
play

An Introduction to Cooperative Game Theory Maria Serna Fall 2016 - PowerPoint PPT Presentation

Definitions Stability notions Other solution concepts Subclasses An Introduction to Cooperative Game Theory Maria Serna Fall 2016 AGT-MIRI Cooperative Game Theory Definitions Stability notions Other solution concepts Subclasses


  1. Definitions TU characteristic function games Stability notions Outcomes Other solution concepts Imputations Subclasses Characteristic Function Games vs. Partition Function Games In general TU games, the payoff obtained by a coalition depends on the actions chosen by other coalitions these games are also known as partition function games (PFG) Characteristic function games (CFG): the payoff of each coalition only depends on the action of that coalition in such games, each coalition can be identified with the profit it obtains by choosing its best action We focus on characteristic function games, and use the term TU games to refer to such games AGT-MIRI Cooperative Game Theory

  2. Definitions TU characteristic function games Stability notions Outcomes Other solution concepts Imputations Subclasses Buying Ice-Cream Game AGT-MIRI Cooperative Game Theory

  3. Definitions TU characteristic function games Stability notions Outcomes Other solution concepts Imputations Subclasses Buying Ice-Cream Game We have a group of n children, each has some amount of money the i -th child has b i dollars. AGT-MIRI Cooperative Game Theory

  4. Definitions TU characteristic function games Stability notions Outcomes Other solution concepts Imputations Subclasses Buying Ice-Cream Game We have a group of n children, each has some amount of money the i -th child has b i dollars. There are three types of ice-cream tubs for sale: AGT-MIRI Cooperative Game Theory

  5. Definitions TU characteristic function games Stability notions Outcomes Other solution concepts Imputations Subclasses Buying Ice-Cream Game We have a group of n children, each has some amount of money the i -th child has b i dollars. There are three types of ice-cream tubs for sale: AGT-MIRI Cooperative Game Theory

  6. Definitions TU characteristic function games Stability notions Outcomes Other solution concepts Imputations Subclasses Buying Ice-Cream Game We have a group of n children, each has some amount of money the i -th child has b i dollars. There are three types of ice-cream tubs for sale: Type 1 costs $7, contains 500g Type 2 costs $9, contains 750g Type 3 costs $11, contains 1kg AGT-MIRI Cooperative Game Theory

  7. Definitions TU characteristic function games Stability notions Outcomes Other solution concepts Imputations Subclasses Buying Ice-Cream Game We have a group of n children, each has some amount of money the i -th child has b i dollars. There are three types of ice-cream tubs for sale: Type 1 costs $7, contains 500g Type 2 costs $9, contains 750g Type 3 costs $11, contains 1kg The children have utility for ice-cream but do not care about money. AGT-MIRI Cooperative Game Theory

  8. Definitions TU characteristic function games Stability notions Outcomes Other solution concepts Imputations Subclasses Buying Ice-Cream Game We have a group of n children, each has some amount of money the i -th child has b i dollars. There are three types of ice-cream tubs for sale: Type 1 costs $7, contains 500g Type 2 costs $9, contains 750g Type 3 costs $11, contains 1kg The children have utility for ice-cream but do not care about money. The payoff of each group is the maximum quantity of ice-cream the members of the group can buy by pooling their money AGT-MIRI Cooperative Game Theory

  9. Definitions TU characteristic function games Stability notions Outcomes Other solution concepts Imputations Subclasses Buying Ice-Cream Game We have a group of n children, each has some amount of money the i -th child has b i dollars. There are three types of ice-cream tubs for sale: Type 1 costs $7, contains 500g Type 2 costs $9, contains 750g Type 3 costs $11, contains 1kg The children have utility for ice-cream but do not care about money. The payoff of each group is the maximum quantity of ice-cream the members of the group can buy by pooling their money The ice-cream can be shared arbitrarily within the group. AGT-MIRI Cooperative Game Theory

  10. Definitions TU characteristic function games Stability notions Outcomes Other solution concepts Imputations Subclasses Ice-Cream Game: Characteristic Function AGT-MIRI Cooperative Game Theory

  11. Definitions TU characteristic function games Stability notions Outcomes Other solution concepts Imputations Subclasses Ice-Cream Game: Characteristic Function Charlie: $6 Marcie: $4 Pattie: $3 AGT-MIRI Cooperative Game Theory

  12. Definitions TU characteristic function games Stability notions Outcomes Other solution concepts Imputations Subclasses Ice-Cream Game: Characteristic Function Charlie: $6 Marcie: $4 Pattie: $3 w = 500 w = 750 w = 100 p = $7 p = $9 p = $11 AGT-MIRI Cooperative Game Theory

  13. Definitions TU characteristic function games Stability notions Outcomes Other solution concepts Imputations Subclasses Ice-Cream Game: Characteristic Function Charlie: $6 Marcie: $4 Pattie: $3 w = 500 w = 750 w = 100 p = $7 p = $9 p = $11 v ( ∅ ) = v ( { C } ) = v ( { M } ) = v ( { P } ) = 0 v ( { C , M } ) = 750 , v ( { C , P } ) = 750 , v ( { M , P } ) = 500 v ( { C , M , P } ) = 1000 AGT-MIRI Cooperative Game Theory

  14. Definitions TU characteristic function games Stability notions Outcomes Other solution concepts Imputations Subclasses Outcomes An outcome of a game Γ = ( N , v ) is a pair ( CS , x ), where: AGT-MIRI Cooperative Game Theory

  15. Definitions TU characteristic function games Stability notions Outcomes Other solution concepts Imputations Subclasses Outcomes An outcome of a game Γ = ( N , v ) is a pair ( CS , x ), where: CS = ( C 1 , ..., C k ) is a coalition structure, i.e., partition of N into coalitions: ∪ k i =1 C i = N C i ∩ C j = ∅ , for i � = j AGT-MIRI Cooperative Game Theory

  16. Definitions TU characteristic function games Stability notions Outcomes Other solution concepts Imputations Subclasses Outcomes An outcome of a game Γ = ( N , v ) is a pair ( CS , x ), where: CS = ( C 1 , ..., C k ) is a coalition structure, i.e., partition of N into coalitions: ∪ k i =1 C i = N C i ∩ C j = ∅ , for i � = j x = ( x 1 , ..., x n ) is a payoff vector, which distributes the value of each coalition in CS: x i ≥ 0, for all i ∈ N � i ∈ C x i = v ( C ), for each C ∈ CS , AGT-MIRI Cooperative Game Theory

  17. Definitions TU characteristic function games Stability notions Outcomes Other solution concepts Imputations Subclasses Outcomes An outcome of a game Γ = ( N , v ) is a pair ( CS , x ), where: CS = ( C 1 , ..., C k ) is a coalition structure, i.e., partition of N into coalitions: ∪ k i =1 C i = N C i ∩ C j = ∅ , for i � = j x = ( x 1 , ..., x n ) is a payoff vector, which distributes the value of each coalition in CS: x i ≥ 0, for all i ∈ N � i ∈ C x i = v ( C ), for each C ∈ CS , feasibility AGT-MIRI Cooperative Game Theory

  18. Definitions TU characteristic function games Stability notions Outcomes Other solution concepts Imputations Subclasses Outcome:example Suppose v ( { 1 , 2 , 3 } ) = 9 and v ( { 4 , 5 } ) = 4 AGT-MIRI Cooperative Game Theory

  19. Definitions TU characteristic function games Stability notions Outcomes Other solution concepts Imputations Subclasses Outcome:example Suppose v ( { 1 , 2 , 3 } ) = 9 and v ( { 4 , 5 } ) = 4 (( { 1 , 2 , 3 } , { 4 , 5 } ) , (3 , 3 , 3 , 3 , 1)) is an outcome AGT-MIRI Cooperative Game Theory

  20. Definitions TU characteristic function games Stability notions Outcomes Other solution concepts Imputations Subclasses Outcome:example Suppose v ( { 1 , 2 , 3 } ) = 9 and v ( { 4 , 5 } ) = 4 (( { 1 , 2 , 3 } , { 4 , 5 } ) , (3 , 3 , 3 , 3 , 1)) is an outcome (( { 1 , 2 , 3 } , { 4 , 5 } ) , (2 , 3 , 2 , 3 , 3)) is NOT an outcome as transfers between coalitions are not allowed AGT-MIRI Cooperative Game Theory

  21. Definitions TU characteristic function games Stability notions Outcomes Other solution concepts Imputations Subclasses Imputations AGT-MIRI Cooperative Game Theory

  22. Definitions TU characteristic function games Stability notions Outcomes Other solution concepts Imputations Subclasses Imputations An outcome ( CS , x ) is called an imputation if it satisfies individual rationality: x i ≥ v ( { i } ) , for all i ∈ N . AGT-MIRI Cooperative Game Theory

  23. Definitions TU characteristic function games Stability notions Outcomes Other solution concepts Imputations Subclasses Imputations An outcome ( CS , x ) is called an imputation if it satisfies individual rationality: x i ≥ v ( { i } ) , for all i ∈ N . Notation: we denote � i ∈ C x i by x ( C ) AGT-MIRI Cooperative Game Theory

  24. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses 1 Definitions 2 Stability notions 3 Other solution concepts 4 Subclasses AGT-MIRI Cooperative Game Theory

  25. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses What Is a Good Outcome? AGT-MIRI Cooperative Game Theory

  26. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses What Is a Good Outcome? The solutions of a game should provide good outcomes. AGT-MIRI Cooperative Game Theory

  27. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses What Is a Good Outcome? The solutions of a game should provide good outcomes. Let us present some stability notions related to outcomes or imputations. AGT-MIRI Cooperative Game Theory

  28. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses What Is a Good Outcome? The solutions of a game should provide good outcomes. Let us present some stability notions related to outcomes or imputations. To simplify the presentation we consider superadditive games. AGT-MIRI Cooperative Game Theory

  29. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Superadditive Games A game G = ( N , v ) is called superadditive if v ( C ∪ D ) ≥ v ( C ) + v ( D ) , for any two disjoint coalitions C and D AGT-MIRI Cooperative Game Theory

  30. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Superadditive Games A game G = ( N , v ) is called superadditive if v ( C ∪ D ) ≥ v ( C ) + v ( D ) , for any two disjoint coalitions C and D Example: v ( C ) = | C | 2 v ( C ∪ D ) = ( | C | + | D | ) 2 ≥ | C | 2 + | D | 2 = v ( C ) + v ( D ) AGT-MIRI Cooperative Game Theory

  31. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Superadditive Games AGT-MIRI Cooperative Game Theory

  32. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Superadditive Games In superadditive games, two coalitions can always merge without losing money; hence, we can assume that players form the grand coalition AGT-MIRI Cooperative Game Theory

  33. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Superadditive Games In superadditive games, two coalitions can always merge without losing money; hence, we can assume that players form the grand coalition Convention: in superadditive games, we identify outcomes with payoff vectors for the grand coalition AGT-MIRI Cooperative Game Theory

  34. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Superadditive Games In superadditive games, two coalitions can always merge without losing money; hence, we can assume that players form the grand coalition Convention: in superadditive games, we identify outcomes with payoff vectors for the grand coalition i.e., an outcome is a vector x = ( x 1 , ..., x n ) with x ( N ) = v ( N ) AGT-MIRI Cooperative Game Theory

  35. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses What Is a Good Outcome? Charlie: $4 Marcie: $3 Pattie: $3 Ice-cream pots: w = (500 , 750 , 100) and p = ($7 , $9 , $11) AGT-MIRI Cooperative Game Theory

  36. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses What Is a Good Outcome? Charlie: $4 Marcie: $3 Pattie: $3 Ice-cream pots: w = (500 , 750 , 100) and p = ($7 , $9 , $11) v ( ∅ ) = v ( { C } ) = v ( { M } ) = v ( { P } ) = 0 v ( { C , M } ) = 500 , v ( { C , P } ) = 500 , v ( { M , P } ) = 0 v ( { C , M , P } ) = 750 AGT-MIRI Cooperative Game Theory

  37. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses What Is a Good Outcome? Charlie: $4 Marcie: $3 Pattie: $3 Ice-cream pots: w = (500 , 750 , 100) and p = ($7 , $9 , $11) v ( ∅ ) = v ( { C } ) = v ( { M } ) = v ( { P } ) = 0 v ( { C , M } ) = 500 , v ( { C , P } ) = 500 , v ( { M , P } ) = 0 v ( { C , M , P } ) = 750 This is a superadditive game, so outcomes are payoff vectors! AGT-MIRI Cooperative Game Theory

  38. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses What Is a Good Outcome? Charlie: $4 Marcie: $3 Pattie: $3 Ice-cream pots: w = (500 , 750 , 100) and p = ($7 , $9 , $11) v ( ∅ ) = v ( { C } ) = v ( { M } ) = v ( { P } ) = 0 v ( { C , M } ) = 500 , v ( { C , P } ) = 500 , v ( { M , P } ) = 0 v ( { C , M , P } ) = 750 This is a superadditive game, so outcomes are payoff vectors! How should the players share the ice-cream? AGT-MIRI Cooperative Game Theory

  39. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses What Is a Good Outcome? Charlie: $4 Marcie: $3 Pattie: $3 Ice-cream pots: w = (500 , 750 , 100) and p = ($7 , $9 , $11) v ( ∅ ) = v ( { C } ) = v ( { M } ) = v ( { P } ) = 0 v ( { C , M } ) = 500 , v ( { C , P } ) = 500 , v ( { M , P } ) = 0 v ( { C , M , P } ) = 750 This is a superadditive game, so outcomes are payoff vectors! How should the players share the ice-cream? (200 , 200 , 350)? AGT-MIRI Cooperative Game Theory

  40. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses What Is a Good Outcome? Charlie: $4 Marcie: $3 Pattie: $3 Ice-cream pots: w = (500 , 750 , 100) and p = ($7 , $9 , $11) v ( ∅ ) = v ( { C } ) = v ( { M } ) = v ( { P } ) = 0 v ( { C , M } ) = 500 , v ( { C , P } ) = 500 , v ( { M , P } ) = 0 v ( { C , M , P } ) = 750 This is a superadditive game, so outcomes are payoff vectors! How should the players share the ice-cream? (200 , 200 , 350)? Charlie and Marcie can get more ice-cream by buying a 500g tub on their own, and splitting it equally AGT-MIRI Cooperative Game Theory

  41. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses What Is a Good Outcome? Charlie: $4 Marcie: $3 Pattie: $3 Ice-cream pots: w = (500 , 750 , 100) and p = ($7 , $9 , $11) v ( ∅ ) = v ( { C } ) = v ( { M } ) = v ( { P } ) = 0 v ( { C , M } ) = 500 , v ( { C , P } ) = 500 , v ( { M , P } ) = 0 v ( { C , M , P } ) = 750 This is a superadditive game, so outcomes are payoff vectors! How should the players share the ice-cream? (200 , 200 , 350)? Charlie and Marcie can get more ice-cream by buying a 500g tub on their own, and splitting it equally (200, 200, 350) is not stable! AGT-MIRI Cooperative Game Theory

  42. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses The core AGT-MIRI Cooperative Game Theory

  43. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses The core The core of a game Γ is the set of all stable outcomes, i.e., outcomes that no coalition wants to deviate from AGT-MIRI Cooperative Game Theory

  44. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses The core The core of a game Γ is the set of all stable outcomes, i.e., outcomes that no coalition wants to deviate from core(Γ) = { ( CS , x ) | x ( C ) ≥ v ( C ) for any C ⊆ N } AGT-MIRI Cooperative Game Theory

  45. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses The core The core of a game Γ is the set of all stable outcomes, i.e., outcomes that no coalition wants to deviate from core(Γ) = { ( CS , x ) | x ( C ) ≥ v ( C ) for any C ⊆ N } each coalition earns at least as much as it can make on its own. AGT-MIRI Cooperative Game Theory

  46. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses The core The core of a game Γ is the set of all stable outcomes, i.e., outcomes that no coalition wants to deviate from core(Γ) = { ( CS , x ) | x ( C ) ≥ v ( C ) for any C ⊆ N } each coalition earns at least as much as it can make on its own. Example: v ( { 1 , 2 , 3 } ) = 9, v ( { 4 , 5 } ) = 4, v ( { 2 , 4 } ) = 7 AGT-MIRI Cooperative Game Theory

  47. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses The core The core of a game Γ is the set of all stable outcomes, i.e., outcomes that no coalition wants to deviate from core(Γ) = { ( CS , x ) | x ( C ) ≥ v ( C ) for any C ⊆ N } each coalition earns at least as much as it can make on its own. Example: v ( { 1 , 2 , 3 } ) = 9, v ( { 4 , 5 } ) = 4, v ( { 2 , 4 } ) = 7 (( { 1 , 2 , 3 } , { 4 , 5 } ) , (3 , 3 , 3 , 3 , 1)) is NOT in the core AGT-MIRI Cooperative Game Theory

  48. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses The core The core of a game Γ is the set of all stable outcomes, i.e., outcomes that no coalition wants to deviate from core(Γ) = { ( CS , x ) | x ( C ) ≥ v ( C ) for any C ⊆ N } each coalition earns at least as much as it can make on its own. Example: v ( { 1 , 2 , 3 } ) = 9, v ( { 4 , 5 } ) = 4, v ( { 2 , 4 } ) = 7 (( { 1 , 2 , 3 } , { 4 , 5 } ) , (3 , 3 , 3 , 3 , 1)) is NOT in the core as x ( { 2 , 4 } ) = 6 and v ( { 2 , 4 } ) = 7 AGT-MIRI Cooperative Game Theory

  49. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses The core The core of a game Γ is the set of all stable outcomes, i.e., outcomes that no coalition wants to deviate from core(Γ) = { ( CS , x ) | x ( C ) ≥ v ( C ) for any C ⊆ N } each coalition earns at least as much as it can make on its own. Example: v ( { 1 , 2 , 3 } ) = 9, v ( { 4 , 5 } ) = 4, v ( { 2 , 4 } ) = 7 (( { 1 , 2 , 3 } , { 4 , 5 } ) , (3 , 3 , 3 , 3 , 1)) is NOT in the core as x ( { 2 , 4 } ) = 6 and v ( { 2 , 4 } ) = 7 no subgroup of players can deviate so that each member of the subgroup gets more. AGT-MIRI Cooperative Game Theory

  50. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Ice-cream game: Core Charlie: $4 Marcie: $3 Pattie: $3 Ice-cream pots: w = (500 , 750 , 100) and p = ($7 , $9 , $11) v ( ∅ ) = v ( { C } ) = v ( { M } ) = v ( { P } ) = 0 v ( { C , M } ) = 500 , v ( { C , P } ) = 500 , v ( { M , P } ) = 0 v ( { C , M , P } ) = 750 AGT-MIRI Cooperative Game Theory

  51. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Ice-cream game: Core Charlie: $4 Marcie: $3 Pattie: $3 Ice-cream pots: w = (500 , 750 , 100) and p = ($7 , $9 , $11) v ( ∅ ) = v ( { C } ) = v ( { M } ) = v ( { P } ) = 0 v ( { C , M } ) = 500 , v ( { C , P } ) = 500 , v ( { M , P } ) = 0 v ( { C , M , P } ) = 750 (200 , 200 , 350) AGT-MIRI Cooperative Game Theory

  52. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Ice-cream game: Core Charlie: $4 Marcie: $3 Pattie: $3 Ice-cream pots: w = (500 , 750 , 100) and p = ($7 , $9 , $11) v ( ∅ ) = v ( { C } ) = v ( { M } ) = v ( { P } ) = 0 v ( { C , M } ) = 500 , v ( { C , P } ) = 500 , v ( { M , P } ) = 0 v ( { C , M , P } ) = 750 (200 , 200 , 350) is not in the core: v ( { C , M } ) > x ( { C , M } ) AGT-MIRI Cooperative Game Theory

  53. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Ice-cream game: Core Charlie: $4 Marcie: $3 Pattie: $3 Ice-cream pots: w = (500 , 750 , 100) and p = ($7 , $9 , $11) v ( ∅ ) = v ( { C } ) = v ( { M } ) = v ( { P } ) = 0 v ( { C , M } ) = 500 , v ( { C , P } ) = 500 , v ( { M , P } ) = 0 v ( { C , M , P } ) = 750 (200 , 200 , 350) is not in the core: v ( { C , M } ) > x ( { C , M } ) (250 , 250 , 250) AGT-MIRI Cooperative Game Theory

  54. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Ice-cream game: Core Charlie: $4 Marcie: $3 Pattie: $3 Ice-cream pots: w = (500 , 750 , 100) and p = ($7 , $9 , $11) v ( ∅ ) = v ( { C } ) = v ( { M } ) = v ( { P } ) = 0 v ( { C , M } ) = 500 , v ( { C , P } ) = 500 , v ( { M , P } ) = 0 v ( { C , M , P } ) = 750 (200 , 200 , 350) is not in the core: v ( { C , M } ) > x ( { C , M } ) (250 , 250 , 250) is in the core: alone or in pairs do not get more. (750, 0, 0) AGT-MIRI Cooperative Game Theory

  55. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Ice-cream game: Core Charlie: $4 Marcie: $3 Pattie: $3 Ice-cream pots: w = (500 , 750 , 100) and p = ($7 , $9 , $11) v ( ∅ ) = v ( { C } ) = v ( { M } ) = v ( { P } ) = 0 v ( { C , M } ) = 500 , v ( { C , P } ) = 500 , v ( { M , P } ) = 0 v ( { C , M , P } ) = 750 (200 , 200 , 350) is not in the core: v ( { C , M } ) > x ( { C , M } ) (250 , 250 , 250) is in the core: alone or in pairs do not get more. (750, 0, 0) is also in the core: AGT-MIRI Cooperative Game Theory

  56. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Ice-cream game: Core Charlie: $4 Marcie: $3 Pattie: $3 Ice-cream pots: w = (500 , 750 , 100) and p = ($7 , $9 , $11) v ( ∅ ) = v ( { C } ) = v ( { M } ) = v ( { P } ) = 0 v ( { C , M } ) = 500 , v ( { C , P } ) = 500 , v ( { M , P } ) = 0 v ( { C , M , P } ) = 750 (200 , 200 , 350) is not in the core: v ( { C , M } ) > x ( { C , M } ) (250 , 250 , 250) is in the core: alone or in pairs do not get more. (750, 0, 0) is also in the core: Marcie and Pattie cannot get more on their own! AGT-MIRI Cooperative Game Theory

  57. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Games with empty core? Let Γ = ( N , v ), where N = { 1 , 2 , 3 } and v ( C ) = 1 if | C | > 1 and v ( C ) = 0 otherwise. AGT-MIRI Cooperative Game Theory

  58. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Games with empty core? Let Γ = ( N , v ), where N = { 1 , 2 , 3 } and v ( C ) = 1 if | C | > 1 and v ( C ) = 0 otherwise. Consider an outcome ( CS , x ). AGT-MIRI Cooperative Game Theory

  59. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Games with empty core? Let Γ = ( N , v ), where N = { 1 , 2 , 3 } and v ( C ) = 1 if | C | > 1 and v ( C ) = 0 otherwise. Consider an outcome ( CS , x ). We have x 1 , x 2 , x 3 ≥ 0, x 1 + x 2 + x 3 = 1, and x i + x j = 1, for i � = j As, x 1 + x 2 + x 3 ≥ 1, for some i ∈ { 1 , 2 , 3 } , x i ≥ 1 / 3. Assume that i = 1, we have x 2 + x 3 = 1 − x 1 ≤ 1 − 1 / 3 ≤ 1! AGT-MIRI Cooperative Game Theory

  60. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Games with empty core? Let Γ = ( N , v ), where N = { 1 , 2 , 3 } and v ( C ) = 1 if | C | > 1 and v ( C ) = 0 otherwise. Consider an outcome ( CS , x ). We have x 1 , x 2 , x 3 ≥ 0, x 1 + x 2 + x 3 = 1, and x i + x j = 1, for i � = j As, x 1 + x 2 + x 3 ≥ 1, for some i ∈ { 1 , 2 , 3 } , x i ≥ 1 / 3. Assume that i = 1, we have x 2 + x 3 = 1 − x 1 ≤ 1 − 1 / 3 ≤ 1! Thus the core of Γ is empty. AGT-MIRI Cooperative Game Theory

  61. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Core on payoff vectors AGT-MIRI Cooperative Game Theory

  62. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Core on payoff vectors Suppose the game is not necessarily superadditive, but the outcomes are defined as payoff vectors for the grand coalition. AGT-MIRI Cooperative Game Theory

  63. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Core on payoff vectors Suppose the game is not necessarily superadditive, but the outcomes are defined as payoff vectors for the grand coalition. Then the core may be empty, even if according to the standard definition it is not. AGT-MIRI Cooperative Game Theory

  64. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Core on payoff vectors Suppose the game is not necessarily superadditive, but the outcomes are defined as payoff vectors for the grand coalition. Then the core may be empty, even if according to the standard definition it is not. Γ = ( N , v ) with N = { 1 , 2 , 3 , 4 } and v ( C ) = 1 if | C | > 1 and v ( C ) = 0 otherwise AGT-MIRI Cooperative Game Theory

  65. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Core on payoff vectors Suppose the game is not necessarily superadditive, but the outcomes are defined as payoff vectors for the grand coalition. Then the core may be empty, even if according to the standard definition it is not. Γ = ( N , v ) with N = { 1 , 2 , 3 , 4 } and v ( C ) = 1 if | C | > 1 and v ( C ) = 0 otherwise not superadditive: v ( { 1 , 2 } ) + v ( { 3 , 4 } ) = 2 > v ( { 1 , 2 , 3 , 4 } ) AGT-MIRI Cooperative Game Theory

  66. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Core on payoff vectors Suppose the game is not necessarily superadditive, but the outcomes are defined as payoff vectors for the grand coalition. Then the core may be empty, even if according to the standard definition it is not. Γ = ( N , v ) with N = { 1 , 2 , 3 , 4 } and v ( C ) = 1 if | C | > 1 and v ( C ) = 0 otherwise not superadditive: v ( { 1 , 2 } ) + v ( { 3 , 4 } ) = 2 > v ( { 1 , 2 , 3 , 4 } ) no payoff vector for the grand coalition is in the core: either { 1 , 2 } or { 3 , 4 } get less than 1, so can deviate AGT-MIRI Cooperative Game Theory

  67. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Core on payoff vectors Suppose the game is not necessarily superadditive, but the outcomes are defined as payoff vectors for the grand coalition. Then the core may be empty, even if according to the standard definition it is not. Γ = ( N , v ) with N = { 1 , 2 , 3 , 4 } and v ( C ) = 1 if | C | > 1 and v ( C ) = 0 otherwise not superadditive: v ( { 1 , 2 } ) + v ( { 3 , 4 } ) = 2 > v ( { 1 , 2 , 3 , 4 } ) no payoff vector for the grand coalition is in the core: either { 1 , 2 } or { 3 , 4 } get less than 1, so can deviate But (( { 1 , 2 } , { 3 , 4 } ) , (1 / 2 , 1 / 2 , 1 / 2 , 1 / 2)) is in the core AGT-MIRI Cooperative Game Theory

  68. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses ǫ -Core AGT-MIRI Cooperative Game Theory

  69. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses ǫ -Core When the core is empty, we may want to find approximately stable outcomes. AGT-MIRI Cooperative Game Theory

  70. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses ǫ -Core When the core is empty, we may want to find approximately stable outcomes. We need to relax the notion of the core: AGT-MIRI Cooperative Game Theory

  71. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses ǫ -Core When the core is empty, we may want to find approximately stable outcomes. We need to relax the notion of the core: core: ( CS , x ) : x ( C ) ≥ v ( C ), for all C ⊆ N AGT-MIRI Cooperative Game Theory

  72. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses ǫ -Core When the core is empty, we may want to find approximately stable outcomes. We need to relax the notion of the core: core: ( CS , x ) : x ( C ) ≥ v ( C ), for all C ⊆ N ǫ -core: { ( CS , x ) : x ( C ) ≥ v ( C ) − ǫ, for all C ⊆ N } AGT-MIRI Cooperative Game Theory

  73. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses ǫ -Core When the core is empty, we may want to find approximately stable outcomes. We need to relax the notion of the core: core: ( CS , x ) : x ( C ) ≥ v ( C ), for all C ⊆ N ǫ -core: { ( CS , x ) : x ( C ) ≥ v ( C ) − ǫ, for all C ⊆ N } Γ = ( N , v ), N = { 1 , 2 , 3 } and v ( C ) = 1 if | C | > 1 and v ( C ) = 0 otherwise AGT-MIRI Cooperative Game Theory

  74. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses ǫ -Core When the core is empty, we may want to find approximately stable outcomes. We need to relax the notion of the core: core: ( CS , x ) : x ( C ) ≥ v ( C ), for all C ⊆ N ǫ -core: { ( CS , x ) : x ( C ) ≥ v ( C ) − ǫ, for all C ⊆ N } Γ = ( N , v ), N = { 1 , 2 , 3 } and v ( C ) = 1 if | C | > 1 and v ( C ) = 0 otherwise 1 / 3-core is non-empty: (1 / 3 , 1 / 3 , 1 / 3) ∈ 1 / 3-core ǫ -core is empty for any ǫ < 1 / 3: x i ≥ 1 / 3, for some i = 1 , 2 , 3, so x ( N \ { i } ) ≤ 2 / 3, v ( N \ { i } ) = 1 AGT-MIRI Cooperative Game Theory

  75. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Least Core AGT-MIRI Cooperative Game Theory

  76. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Least Core If an outcome ( CS , x ) is in the ǫ -core, the deficit v ( C ) − x ( C ) of any coalition is at most ǫ AGT-MIRI Cooperative Game Theory

  77. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Least Core If an outcome ( CS , x ) is in the ǫ -core, the deficit v ( C ) − x ( C ) of any coalition is at most ǫ We are interested in outcomes that minimize the worst-case deficit AGT-MIRI Cooperative Game Theory

  78. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Least Core If an outcome ( CS , x ) is in the ǫ -core, the deficit v ( C ) − x ( C ) of any coalition is at most ǫ We are interested in outcomes that minimize the worst-case deficit Let ǫ ∗ (Γ) = inf { ǫ | ǫ -core of Γ is not empty } . AGT-MIRI Cooperative Game Theory

  79. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Least Core If an outcome ( CS , x ) is in the ǫ -core, the deficit v ( C ) − x ( C ) of any coalition is at most ǫ We are interested in outcomes that minimize the worst-case deficit Let ǫ ∗ (Γ) = inf { ǫ | ǫ -core of Γ is not empty } . It can be shown that, for all Γ, the ǫ ∗ (Γ)-core is not empty. AGT-MIRI Cooperative Game Theory

  80. Definitions Core and variations Stability notions Fairness: Shapley value Other solution concepts Computational Issues Subclasses Least Core If an outcome ( CS , x ) is in the ǫ -core, the deficit v ( C ) − x ( C ) of any coalition is at most ǫ We are interested in outcomes that minimize the worst-case deficit Let ǫ ∗ (Γ) = inf { ǫ | ǫ -core of Γ is not empty } . It can be shown that, for all Γ, the ǫ ∗ (Γ)-core is not empty. The ǫ ∗ (Γ)-core is called the least core of Γ and ǫ ∗ (Γ) is called the value of the least core AGT-MIRI Cooperative Game Theory

Recommend


More recommend