Support Vector Machine: Detecting Linear Separable Class By Boolean Logic Gates

Abstract

Artificial Neural Network is used as promising tool to resolve various real world problems. Most often the patterns are derived from real world measurement and cluster together based on some properties in certain regions. Pattern recognition can interpret as classification of data based on certain knowledge. The task of classification is to assign a test object to one of two classes. This paper focuses on linear and non-linear separability using classification of dataset. Generally linear separable problems are simple to solve than non-linear separable problem. We considered simple boolean function AND and OR for classification. To carry out the experiment, Zoo data set is used. SVM yields excellent performance on classification problem. We measured accuracy score using SVM