ABSTRACT

In order to measure the profitability of an industry by predicting Pre-Tax Operating Margin by applying regression technique on Price/Sales Ratio and Net Margin of various industries. Prediction of Pre-Tax Operating Margin is done using Support vector Regression (SVR). We present a model in this paper in order to solve the problem of over-fitting which is due to noise and outliers in dataset. For this a weighted coefficient based approach is proposed that reduces the prediction error and provides the higher accuracy than simple support vector regression. At last, the comparison of SVR using different kernel functions with weight is done and results of experiments shows that LS-SVR with RBF kernel function using weighted coefficient have better accuracy.