Shapley Feature Selection

Feature selection is a popular topic. The main approaches to deal with it fall into the three categories of filters, wrappers and embedded methods. Advancement in algorithms, though proving fruitful, may be not enough. We propose integrate an explainable AI approach, based on Shapley values, provide more accurate information for feature selection. test our proposal real setting, which concerns prediction probability default Small Medium Enterprises. Our results show that integrated approach indeed prove fruitful some methods, particular parsimonious ones like LASSO. In general combination seems useful algorithm can improve their performance with.