Paper Title
Efficient Feature Selection Using Sequential Backward Selection Algorithm In The Classification Of Histopathological Images

Abstract
In this study, the histopathological images were classified by using Bayesian classifier with SBS algortihm. Histopathological images were taken from patients with hepatitis and healty persons. Histopathological images were converted gray images. The statistical properties of the gray images were obtained. These features (expected value, standard deviation, skewness, kurtosis, entropy) were given to Bayesian clasifier by using SBS algorithm and effective features were detected. 100 training data and 52 test data were used for the classification. As a result, all of the features were classified by using SBS.Kurtosis and entropy were determined as the most effective features. By using kurtosis and entropy were obtained %98,08accuracy for classification. Index Terms� Sbs, Bayesian, Hepatitis, Histopathological Images, Statistical Feature, Classification, Effective Feature.