A Method for Improving the Accuracy of Predicting Protein Structural Classes

Abstract

The structure of a protein is closely correlated to its function. Feature dimension reduction method is one of most famous machine learning tools. Some researchers have begun to explore feature dimension reduction method for computer vision problems. Few such attempts have been made for classification of high-dimensional protein data sets. In this paper, feature dimension reduction method is employed to reduce the size of the features space. Comparison between linear Feature dimension reduction method and nonlinear feature dimension reduction method is performed to predict protein structural classes. The results with high success rates indicate that the above method is used effectively to deal with this complicated problem of predicting proteins structural classes.