An empirical assessment of different kernel functions on the performance of support vector machines

Artificial intelligence (AI) and machine learning (ML) have influenced every part of our day-to-day activities in this era technological advancement, making a living more comfortable on the earth. Among several AI ML algorithms, support vector (SVM) has become one most generally used algorithms for data mining, prediction other (AI ML) domains. The SVM’s performance is significantly centred kernel function (KF); nonetheless, there no universal accepted ground selecting an optimal KF specific domain. In paper, we investigate empirically different KFs SVM various fields. We illustrated based through extensive experimental results. Our empirical results show that single always suitable achieving high accuracy generalisation all However, gaussian radial basis (RBF) often default choice. Also, if parameters RBF exponential are optimised, they outperform linear sigmoid method terms accuracy. Besides, linearly separable dataset.