Stability of different feature selection methods for selecting protein sequence descriptors in protein solubility classification problem

摘要

Predicting protein solubility has gained lots of intention in the recent years and several descriptors have been defined to describe proteins in these works. Therefore, different feature selection methods have been used for selecting the most important attributes. An empirical study, that aims to explain the relationship between the number of samples and stability of seven different feature selection techniques for protein datasets, is presented.