Power Quality Data Mining Using Improved Probabilistic Neural Network

This paper deals with the Technical Analysis which helps the investors to discover hidden patterns from the historic financial and time series stock market data. These techniques have the probable capability to help the investors in their investment decisions by looking into the new hidden patterns and opportunities. It maximizes the prediction of financial stock market using time series quantitative analysis. Two analytics are devised by combining various financial indicators. A back test on the historic data is performed to calculate the percentage of profitable buy / sell signals generated by the combinational analytics. The Moving Average Crossover (MAC) Algorithm is taken as a benchmark to compare the percentage of profitable buy / sell signals generated by the proposed combinational algorithms. Both the proposed algorithms outperform the MAC algorithm.