Feature Selection Based on the Shapley Value
نویسندگان
چکیده
We present and study the Contribution-Selection algorithm (CSA), a novel algorithm for feature selection. The algorithm is based on the Multiperturbation Shapley Analysis, a framework which relies on game theory to estimate usefulness. The algorithm iteratively estimates the usefulness of features and selects them accordingly, using either forward selection or backward elimination. Empirical comparison with several other existing feature selection methods shows that the backward elimination variant of CSA leads to the most accurate classification results on an array of datasets.
منابع مشابه
Improving LNMF Performance of Facial Expression Recognition via Significant Parts Extraction using Shapley Value
Nonnegative Matrix Factorization (NMF) algorithms have been utilized in a wide range of real applications. NMF is done by several researchers to its part based representation property especially in the facial expression recognition problem. It decomposes a face image into its essential parts (e.g. nose, lips, etc.) but in all previous attempts, it is neglected that all features achieved by NMF ...
متن کاملObtaining a possible allocation in the bankruptcy model using the Shapley value
Data envelopment analysis (DEA) is an effective tool for supporting decision-makers to assess bankruptcy, uncertainty concepts including intervals, and game theory. The bankruptcy problem with the qualitative parameters is an economic problem under uncertainty. Accordingly, we combine the concepts of the DEA game theory and uncertain models as interval linear programming (ILP), which can be app...
متن کاملA Hybrid Grey-Game-MCDM Method for ERP Selecting Based on BSC
An enterprise resource planning (ERP) software is needed for industries and companies that want to develop in future. Many of the manufactures and companies have a problem with ERP software selection. An inappropriate selection process can affect both the implementation and the performance of the company significantly. Although several models are proposed to solve this problem many of them did n...
متن کاملAlternative Axiomatic Characterizations of the Grey Shapley Value
The Shapley value, one of the most common solution concepts of cooperative game theory is defined and axiomatically characterized in different game-theoretic models. Certainly, the Shapley value can be used in interesting sharing cost/reward problems in the Operations Research area such as connection, routing, scheduling, production and inventory situations. In this paper, we focus on the Shapl...
متن کاملInvestigating the Correlation of Selected Banks with Dynamic Conditional Correlation (DCC) Model and Identifying Systemically Important Banks with Conditional Value at Risk and Shapley Value Method
Systemic risk arises from simultaneous movement or correlations between market segments; Thus, systemic risk occurs when there is a high correlation between the risks and crises of different market segments or institutions operating in the economy, or when the risks of different segments in a market segment or a country are related to other segments and other countries. This paper presents a me...
متن کامل