Stock Market Prediction using Financial Cybernetics

Abstract

Research on stock markets is unusual because of random walk performance of stock time series. Forecasting in stock market returns is a significant problem in finance. Currently, the essentials of stock market cannot be isolated from human life because of the frequent investment done in the stock market by people worldwide. While making investments in stock market, traders not only purchase a stock whose value is likely to increase in the future but also, they are likely to refrain from purchasing those stocks whose value is expected to drop in the future. Therefore, a precise estimation of the movement of stock price in the market in order to make maximum profit and curtail loss is urgently required.At large, it is tough to explore specific training algorithm that works best under all conditions at all time therefore, revolutionizing strategies and optimization protocols to develop stochastic methods for training artificial neural networks. A complete study of evolutionary algorithms in developing artificial neural networks can be found. Very recently, an analytical analysis for the stock price movement prediction was reported during the demonetization period in India. Recently, our group has shown the capabilities of artificial neural network in the field of atmospheric physics by the prediction of the ozone hole area. In this paper, comprehensive working of multi-layered neural network along with brief report of various activation function is presented. Using the given information, neural network is skilled, and predictions are stated for Reliance Industries listed in National Stock Exchange, India.

 Keywords: Stock Markets, returns, investment, price.