Stock Price Prediction Using Machine Learning

Stock Price Prediction Using Machine Learning

Ramesh B, Shrishail Telasang, Rekha Jambagi, Pradeep Kumar H M, Savitha M

Asst. Professor, U G Student Dept of CS

City engineering college

Bangalore-560062, Karnatak

rameshb@cityengineeringcollege.ac.in

sktelasang990@gmail.com

jambagirekha@gmail.com

pradeephm224@gmail.com

savithaarun072@gmail.com

Abstract—Forecasting stock market prices is a challenging task for traders, analysts, and engineers due to the myriads of variables influencing stock prices. However, the advent of artificial intelligence (AI) and natural language processing (NLP) has significantly advanced stock market forecasting. AI’s ability to analyze complex data sets allows for more informed predictions. Despite these advancements, stock price forecasting remains an area where AI has not yet achieved optimal results. In this paper, we forecast stock prices using 30 years of historical data from various national banks in India sourced from the National Stock Exchange. We employ advanced deep learning models, including multivariate multi-step Long Short-Term Memory (LSTM), Facebook Prophet with LightGBM and Optuna, and Seasonal Auto-Regressive Integrated Moving Average (SARIMA). Additionally, we analyze news data from tweets and reliable sources like Business Standard and Reuters, recognizing their significant impact on stock price movements.

Index Terms—Stock Market Forecasting, Data preprocess- ing, Advanced Algorithms, Deep Learning Models, Multivari- ate Multistep LSTM, SARIMA, LightGBM, Optuna, acebook Prophet,Sentiment Analysis.