Analysis and prediction of Indian stock market: a machine-learning approach

Prediction of financial stock market is a challenging task because of its volatile and non- linear nature. The presence of different factors like psychological, sentimental state, rational or irrational behaviour of investors make the stock market more dynamic. With the inculcation of algorithms based on artificial intelligence, deep learning algorithms, the prediction of movement of financial stock market is revolutionized in the recent years. The purpose of using these algorithms is to help the investors for taking decisions related to the Stock Pricing. A model has been proposed to predict the direction of movement of Indian stock market in the near future. This model makes use of historical Indian stock data of companies in nifty 50 since they came existence along with some financial and social indicators like financial news and tweets related to stocks. After pre-processing and normalization various machine learning algorithms like LSTM, support vector machines, KNearest neighbour, random forest, gradient boosting regressor are applied on this time series data to produce better accuracy and to minimize the RMSE error. This model has the ability to reduce major losses to the investors who invest in stock market. The social indicators will give an insight for predicting the direction of stock market. The LSTM network will make use of historical closing prices, tweets and trading volume.