Hybrid Classification Algorithms for Predicting Student Performance

Data mining is the progress of instinctively find out valuable information from huge depot. In educational data mining predicting the academic performance of the student is a well-known research. The intention of the research work is to examine the classification algorithms as a hybrid classification. We used Radial Basis Function network, Multilayer Perceptron, C4.5 and Random Forest Algorithms for classification. Initially classification accuracy was computed individually by the classification algorithms. The Radial Basis Function network, Multilayer Perceptron, C4.5 and Random Forest Algorithm’s individual classification gave the accuracy of 72.9167%, 75.4167%, 75% and 73.125% vice versa. To increase more accuracy of classification algorithm the Radial Basis Function network is combined with multilayer perceptron. This hybrid algorithm provides 75.625% of classification accuracy. Then we combined C4.5 algorithm with random forest algorithm which gives 76.4583% classification accuracy. In this study we found hybrid classification algorithm gives more accuracy than individual classification algorithm.