Parameters of support vector machines (SVMs) such as kernel parameters or the penalty parameter have a strong influence on the accuracy and complexity of learnt classification models. In recent work, different evolutionary optimization algorithms were employed for optimizing SVMs; in this paper, we propose a social ski-driver (SSD) optimization algorithm which is inspired by evolutionary optimization algorithms for optimizing parameters of SVMs, with the aim of improving the classification performance. We also propose a variant of SSD for coping with the problem of imbalanced data, which is one of the challenging problems for building robust classification models. In this study, eight standard imbalanced datasets were used for testing the proposed algorithm. Experimental results show that the SSD algorithm is capable of finding near-optimal values of SVMs parameters.

Access to the full paper by A. Tharwat and T. Gabel: "Parameter Optimization of Support Vector Machines for Imbalanced Data Using the Social Ski Driver Algorithm" on SpringerNature