Comparison of Machine Learning Algorithms for Raw Handwritten Digits Recognition

摘要

Handwritten digit recognition has been an area of interest for many years. Most of the published work present the accuracy of the classification algorithms. Accuracy alone should not be used to evaluate the performance of the algorithms. In this work, the machine learning algorithms' accuracy, precision, recall and F1-score are considered for non-neural network based algorithms such as support vector machines, K-nearest neighbor, decision trees and logistic regression. In addition to performance metrics the training and validation times are also presented. This documented information on the classification performance in conjunction with the time required for classification will enable the potential users to choose the particular method wisely for their applications.