A non parametric Partial Histogram Bayes learning algorithm for classification applications

In many applications such as dealing with database, continuous environment and humanoid robots, the machine often deals with large amount of data every day of work. Dealing with large amount of data requires fast as well as accurate learning algorithms to do the classification. A new supervised non parametric Partial Histogram Bayes learning algorithm (PHBayes) is proposed and presented in this paper. The proposed algorithm was tested on image database and compared with other standard algorithms like Naïve Bayes, Gaussian Mixture Model based Classifier, 1^st Nearest Neighbor and Nearest Class Mean for classification purpose. The experimental results showed that the proposed algorithm is faster as well as more accurate compare with other algorithms, which makes it worthy to be considered in classification applications.