A Review of Support Vector Machines for Handwritten Character Recognition

This paper investigates the applications of Support Vector Machines (SVMs) in handwritten character recognition. The previous research works are included and reviewed in three categories: multi-class SVM based approaches, SVM ensemble, and hybrid approaches combining SVMs and other classifiers. The performance of SVMs is compared with that of K-Nearest Neighbor, Multi-Layer Perceptron, and Boosting through experiments on MNIST digit database. Finally, the conclusions are given and possible directions of future research are discussed.