Epilepsy is associated with the abnormal electrical activity in the brain which is detected by recording EEG (Electroencephalogram) signals. This signal is non-linear and chaotic and hence, it is very time-consuming and tedious to analyse them visually. In this work, we have extracted five entropy features such as Approximate Entropy, Sample Entropy, Fuzzy Entropy, Permutation Entropy and Multi-scale Entropy for characterizing the focal signals. We have used Sequential Forward Feature Selection (SFFS) algorithm to select two significant features for epilepsy classification. These two features are given as input to the Least Square Support Vector Machine (LS-SVM) classifier to differentiate normal and focal signal. The classification accuracy of our method is 82%.  Moreover, the average computational time for the selected feature set is 47.94 seconds.