Stock Price Prediction Using LSTM Model

Tasneem Rahath; Duluri Harika, Nuthalaganti Meenakshi, Asfiya Naaz

doi:10.62647/

Authors

Tasneem Rahath Assistant Professor, Department Of Information Technology, Bhoj Reddy Engineering College For Women, India Author
Duluri Harika, Nuthalaganti Meenakshi, Asfiya Naaz B. Tech Students, Department Of Information Technology, Bhoj Reddy Engineering College For Women, India. Author

DOI:

https://doi.org/10.62647/

Keywords:

LSTM (Long Short-Term Memory), Time Series Forecasting, Machine Learning, Historical Data, Stock Prediction, Price.

Abstract

Stock price prediction is a critical yet complex task in the financial sector due to the volatile and dynamic nature of the market. Traditional prediction methods like statistical analysis and basic machine learning often fail to capture long-term dependencies and real-world market influences. To address these limitations, this project proposes a web-based stock price prediction system using Long Short-Term Memory (LSTM) networks—a type of Recurrent Neural Network (RNN) designed to model sequential data effectively.

The system is built to analyze historical stock data and generate accurate future price forecasts. It features a user-friendly web interface developed with React and a backend powered by Flask and TensorFlow/Keras. Key functionalities include data collection, preprocessing, model training, real-time prediction, and graphical visualization using Chart.js or Plotly.js. The model is trained using CSV data and enhanced through feature selection techniques like PCA and RFE.

This solution offers improved prediction accuracy, scalability to multiple stocks, and automated analysis, making it a valuable tool for investors and financial analysts. The project is developed as a mini-project under the Information Technology curriculum and fulfills the partial requirements for the award of the Bachelor of Technology degree.