Abstract:Shapley value is a concept in cooperative game theory for measuring the contribution of each participant, which was named in honor of Lloyd Shapley. Shapley value has been recently applied in data marketplaces for compensation allocation based on their contribution to the models. Shapley value is the only value division scheme used for compensation allocation that meets three desirable criteria: group rationality, fairness, and additivity. In cooperative game theory, the marginal contribution of each contributor to each coalition is a nonnegative value. However, in machine learning model training, the marginal contribution of each contributor (data tuple) to each coalition (a set of data tuples) can be a negative value, i.e., the accuracy of the model trained by a dataset with an additional data tuple can be lower than the accuracy of the model trained by the dataset only. In this paper, we investigate the problem of how to handle the negative marginal contribution when computing Shapley value. We explore three philosophies: 1) taking the original value (Original Shapley Value); 2) taking the larger of the original value and zero (Zero Shapley Value); and 3) taking the absolute value of the original value (Absolute Shapley Value). Experiments on Iris dataset demonstrate that the definition of Absolute Shapley Value significantly outperforms the other two definitions in terms of evaluating data importance (the contribution of each data tuple to the trained model).

A simplified expression of the Shapley function for fuzzy game

Applying the model of game with fuzzy coalition to the optimum configuration of cooperation supply chain

Absolute Shapley Value

Fuzzy modeling incorporated with fuzzy d-s theory and fuzzy naive bayes

Cooperative networks and f-Shapley value

Shapley Value Approximation Based on Complementary Contribution

PN Equilibrium Strategy for Matrix Games with Fuzzy Payoffs

Fuzzy Harsanyi solutions for fuzzy level structure games with multi weight systems

Improved Shapley Values Based on Players’ Least Square Contributions and Their Applications in the Collaborative Profit Sharing of the Rural E-commerce

The Shapley Value of Cooperative Game with Stochastic Payoffs

Shapley Meets Shapley

The Shapley-Hodge Associated Game

Approximating the Shapley Value without Marginal Contributions

Linearly representable games and pseudo-polynomial calculation of the Shapley value

Cooperative game theory based consistency and consensus reaching process for group decision making with fuzzy preference relations

Generalized fuzzy rough sets based on fuzzy coverings

The interrelation in fuzzy clustering

Some T-Spherical Hesitant Fuzzy Shapley Bonferroni Mean Operators and Their Applications

Efficient Sampling Approaches to Shapley Value Approximation

Games in possibility capacities with payoff expressed by fuzzy integral

Ambitropical geometry, hyperconvexity and zero-sum games