Stock Prediction Using Machine Learning and Sentiment Analysis

Predicting stock market behavior has been a sector of interest to many researchers, especially in the field of statistics and data analysis. The ability to analyze the stock market that appears to lack consistency, but also appears to be impacted by historical events is still a challenge. Long Short-Term Memory (LSTM) network model, Linear Regression model, Autoregressive Integrated Moving Average (ARIMA), and the stock market prediction using Sentiment Analysis are included in the proposed model. The dataset used is based on real-time stock market data from Yahoo finance website. The calculated root mean square value predicts how much the data is concentrated near the best equity line. Emotional/Sentiment Analysis is done using twitter API and natural language processing. The result obtained proves that the proposed model can be used to predict future stock market behavior.