A QSAR classification model of skin sensitization potential based on improving binary crow search algorithm

Classifying of skin sensitization using the quantitative structure-activity relationship (QSAR) model is important. Applying descriptor selection is essential to improve the performance of the classification task. Recently, a binary crow search algorithm (BCSA) was proposed, which has been successfully applied to solve variable selection. In this work, a new time-varying transfer function is proposed to improve the exploration and exploitation capability of the BCSA in selecting the most relevant descriptors in QSAR classification model with high classification accuracy and short computing time. The results demonstrated that the proposed method is reliable and can reasonably separate the compounds according to sensitizers or non-sensitizers with high classification accuracy.