Foreseeing Stock Price- A Novel Modelling Approach Using Artificial Intelligence

Abstract

Predicting stock market is always a challenging work. As more and more people are involved in trading , its always a optimistic behaviour to know the future stock price. By knowing the future stock price, investors and traders can sell or buy an appropriate stock. In general to predict the stock price, investors  used fundamental analysis, technical analysis and technological analysis. Other methods such as using stock indicators also proved to be a worth some. But whatever the methods used, the impact of news about the particular stock is always a determining factor for finding its stock price. Therefore, the investors will have to keep a track on the news channel so that they can estimate their profit level considerably. Most of the previous work in the industry focussed only on the news about the stock and not on their impact. In this work we proposed a novel methodology that integrates mathematical functions, indicators and news information by implementing a  soft computing approach. Particularly we are aiming to provide the next day stock price by the use of fuzzy logic, neural network and genetic algorithm.In this paper we have applied sentiment analysis and soft computing techniques together  develop a optimised classifier F news extracted from financial sectors and media information. The  improved model can give useful result for the investors in finding the stock price.