MCA (School of Computer Science)
MINOR PROJECT
ABSTRACT
Stock Price Prediction of IT Midcaps Companies using ML Models
Motivation: Stock price movements are influenced by a variety of financial and market factors. For Indian IT midcap
companies, understanding how past quarterly results impact share prices can provide valuable insights for investors,
analysts, and financial institutions. This research aims to analyse the relationship between financial metrics and stock
price fluctuations, enabling better decision-making in the investment domain.
Tools and Technology:
• Web Development – HTML, CSS, and JavaScript for an interactive interface to input match details.
• Machine Learning – Scikit-learn for training predictive models.
• Data Processing – Pandas, NumPy, and Matplotlib for data cleaning, feature engineering, and visualization.
Proposed Model: The study follows a structured approach to stock price prediction. First, correlation analysis is
conducted to identify key financial metrics that influence stock prices. Next, Multiple Linear Regression is employed to
quantify the impact of variables such as Market Cap, Price-to-Earnings (P/E) Ratio, Face Value, Net Profit, and Earnings
Per Share (EPS) on share prices. Additionally, a Random Forest model is used to enhance prediction accuracy and
account for non-linear relationships. The performance of both models is compared to assess their effectiveness.
Application Areas: The findings of this research have broad implications across various domains, including financial
analytics, investment strategy formulation, and stock market forecasting. Investors can leverage these insights to make
informed decisions, while financial analysts can refine their predictive models for better market assessments.
Furthermore, fintech companies can integrate such models into their platforms to enhance stock recommendation
systems.
Conclusion: Our study highlights the significant impact of select financial metrics on the stock prices of Indian IT
midcap companies. While certain parameters like Market Cap and EPS exhibit strong predictive power, others such as
Sales and Operating Profit show limited influence. The discrepancies in prediction accuracy across companies
emphasize the role of data quality, feature selection, and market dynamics. Future research can focus on integrating
more advanced machine learning models and external factors such as macroeconomic indicators to improve prediction
accuracy.
References:
[1] S. K. Pandey, V. Kumar, and R. Pandey, “Stock market prediction: A review of literature and future directions,” International Journal of
Computer Science and Engineering, vol. 7, no. 6, pp. 82–91, 2017.
[2] A. Vij, K. Saxena, and A. Rana, “Stock market prediction using machine learning techniques,” AIIT, Amity University Uttar Pradesh,
Noida, India.
[3] S. Alex, S. Purakayastha, S. Chattaraj, and A. Jadhav, “Stock market prediction using deep learning approaches,” in 2024 Second
International Conference on Emerging Trends in Information Technology and Engineering (ICETITE), 2024.
[4] X. Jiang, “Deep learning models for stock market prediction: A review,” Neural Computing, vol. 32, no. 3, pp. 547–568, 2021.
Submitted by: Internal Guide
Soumyajit Singh Mr. Manjit Kumar Nayak
Regd No: 24240027 (Ast.Prof, School of computer Science)
