Prediction of Market trends using Machine Learning

Abstract

Forecasting about Indian market has always been interesting and topic of discussion among analyst and researchers. With the arrival of machine learning and artificial intelligence the race is now becoming the competition with best algorithms to be used and give investors more profit. In past years prediction was only based on experience and daily headlines of business newspapers but now it depends on various international, national and political economic factors and the sentiments and reaction of people over the issues. With the growing power of social media, the game in market over this also changed now with the help of sentiment analysis over social media we can determine the mood of investors over the news. In the present scenarios you can divide two categories for the prediction strategies one is the time series analysis of stocks and the second is artificial intelligence property over the market. AI property contains multi-layer perception, SVM, naive Bayes , back propagation,CNN,LSTM, RNN etc during this we have a tendency to came with plan of combination of each. Within the paper we have conjointly covers the assorted challenges that are encountered while building prediction models. This whole module focuses on use of statistical analysis and conjointly development of the sentimental analysis and to get better results. The LSTM has the advantage of analyzing relationship between time series knowledge through memory functions. The performance of the system is improved by combine efforts of time-series and sentiments with the LSTM prediction model.