An Improved Binary Classification Framework for Investment Class Rating

This paper presents a binary classification scheme for investment class rating using support vector machine (SVM). The suggested SVM model is trained offline and takes twelve financial ratios as attributes from different standard investment companies as inputs and correctly classify whether it is a good investment grade or bad investment grade company as output. The overall performance of SVM strongly depends on the regularization parameter C and kernel parameter ?. Hence, we propose the PSO based optimization technique using mean square error (MSE) as the fitness function to optimize the value of C and ?. The proposed scheme is implemented using Matlab and Libsvm tool. Comparison is made in terms of different performance measures like classification accuracy, sensitivity, specificity, precision, confusion matrix etc. From experimental results and analysis, it is observed that the proposed scheme has a superior performance as compared to SVM based approach without parameter optimization and neural network based scheme.