Improving amphetamine-type stimulants drug classification using chaotic-based time-varying binary whale optimization algorithm

• Main Authors: Draman @ Muda, Azah Kamilah, Mohd Yusof, Norfadzlia, Pratama, Satrya Fajri, Carbo-Dorca, Ramon, Abraham, Ajith
• Format: Article
• Language: English
• Published: Elsevier B.V. 2022
• Online Access: http://eprints.utem.edu.my/id/eprint/26392/2/2022_CILS_ONLINE-VERSION_NORFADZLIA_PAPER2.PDF; http://eprints.utem.edu.my/id/eprint/26392/; https://www.sciencedirect.com/science/article/pii/S0169743922001460
• Institution: Universiti Teknikal Malaysia Melaka

A new chaotic time-varying binary whale optimization algorithm (CBWOATV) is introduced in this paper to optimize the feature selection process in Amphetamine-type Stimulants (ATS) and non-ATS drugs classification. Two enhancement methods were introduced in this study to provide a fit balance between exploration and exploitation in standard WOA. Firstly, a non-linear time-varying modified Sigmoid transfer function is used as the binarization method. Second, a hybrid Logistic-Tent chaotic map is employed to substitute the pseudorandom numbers of the probability operator in standard WOA. Specific high-dimensional molecular descriptors of ATS and non-ATS drugs were employed to evaluate the efficiency of the proposed algorithm. Experimental results and statistical analysis indicate that the proposed CBWOATV algorithm can prevent the problem of stagnation and entrapment in local minima in WOA. As a result, optimal descriptors were selected contributing to enhanced classification performance.