The International Journal of

The problem addressed in this research is to evaluate the two protein secondary structure methods (neural networks and support vector machines) and then make a comparison among both the methods. From the experimental results which are obtained from the quantitative comparison of Neural Network and Support Vector Machine for Protein Secondary Structure Prediction we can conclude that, when given with same training and test data, Support Vector Machine performs slightly better than Neural Network. The difference among prediction accuracies of both the methods can be considered significant due to the nature of progress in the area of Protein-SSP, i.e. in general the accuracy more than 70% is considered very significant. Furthermore, the SVM method has its unique merits and it can successfully avoid many problems which other machine learning approaches often encounter. SVM is a promising method in practice. Further research to improve the accuracy of protein secondary structure prediction is needed.