PREDICTION OF FUTURE STOCK PRICES EXPLOITING MACHINE LEARNING TECHNIQUES

Abstract

newline newline newlineStock price prediction is the significant task in financial decision making and investing Since the stock values changes continuously it is very challenging to forecast the future stock value In spite of the possibility of drop in the value of the share because of market fluctuations there still exists risk These fluctuations impact the trading volume and the stock price which makes the prediction even more challenging Thus in general forecasting the behavior of the stock market through a variety of methods and techniques is helpful in assisting investors to react with greater certainty to know when to buy or sell the stock Thus a huge variety of techniques have been introduced to predict the variations in the stock prices However numerous efforts have been made no techniques accurately predicts the stock price movement newlineData mining and evolutionary techniques have been rapidly growing methods for solving the issues in stock price prediction Data mining is the process of mining or extracting huge volume of information from large databases Since the volume of data in stock market is huge the prediction becomes challenging Thus data mining techniques are designed to aid investors in discovering hidden patterns from the historical data that contains feasible predictive capability Automated programs have been developed using the predictive and data mining techniques for the stock markets Since the historical data is vital in developing the prediction models it assists investors in extracting hidden patterns from the historical data and enables the prediction of the future returns newlineOn the other hand evolutionary algorithm plays a vital role in forecasting the stock prices accurately Several researches on the application of evolutionary techniques have shown its benefits over other parametric approaches Evolutionary techniques are easy to implement easy to understand does not experience the ill effects of dimensionality and can be utilized to optimize more formal techniques newlineThus in this research