Artificial Bee Colony Algorithm for Feature Selection in Fraud Detection Process

More and more, nowadays, better performance and quality of current classifiers are required when the topic is fraud detection. In this context, processes such as feature selection help to increase the quality of the results obtained by the existing classifiers in the literature, since the high dimensionality of current datasets and redundant information significantly affect the performance of these techniques. This work proposes a wrapper method of feature selection using the ABC algorithm combined with Logistic Regression classification, seeking to obtain better results for fraud detection. Through the tests performed and the results obtained, it is observed that the reduction in the number of features can reduce the database complexity and achieve a higher accuracy in classification when compared to the set classification when using all its attributes. It is also notable the effectiveness of the method as it reaches the proposed objective with as much as quality as other well-known methods while also contributing to optimizing parameters of other feature selection algorithms.