Recognizing handwritten characters using support vector machine algorithm and compare its feature extraction precision with Naive Bayes algorithm.

The aim is to predict the handwritten characters using algorithms such as Support Vector Machine (SVM) algorithm and compare its feature extraction precision with Naive Bayes (NB) algorithm and compare the accuracies and precisions obtained by both algorithms. In the proposed work, the recognition of the handwritten character was carried out by machine learning algorithms namely NB algorithm (n=30) and SVM algorithm (n=30) and achieved good accuracy by both the algorithms. In the existing system, there is a lack of prediction on handwritten characters recognition with G power 80% and alpha value 0.05. The NB algorithm has higher accuracy than SVM algorithm for recognizing handwritten character recognition. The NB algorithm performed with an accuracy of 97.45% which appears better than the SVM algorithm with accuracy of 93.44%. The significance value of the NB algorith m is 0.037. Within the discussions in this study, it was proved that the NB algorithm gained more accuracy in recognizing the handwritten characters than the SVM algorithm. [ABSTRACT FROM AUTHOR]