HomeJournalsTBFLIVol. 1, Iss. 2Forecasting Stock Prices: A Machine Learning-Based
tbfli
Research ArticleTransactions on Banking, Finance, and Leadership Informatics

Volume 1, Issue 2 · 25 October 2025

ISSN: 3067-5804 · E-ISSN: 3067-5812

Forecasting Stock Prices: A Machine Learning-Based Approach for Predictive Analytics Through a Case Study

Article ID:tbfli_25009

Abstract

Stock price prediction has always been a challenging task, requiring careful observation of trends and dynamics of the market because of the volatile and complex nature of financial markets. Various factors affect market behavior all the time. Even some unquantifiable factors like 25 Oct 2025 (Published Online) emotions of the masses, social and political dynamics, etc., also play a great role. So perfect Machine Learning, Deep Learning, behaviors into consideration is crucial for better prediction of the ups and downs of prices. SMA, EMA, RSI, MACD, Bollinger Various machine learning and deep learning models have been proposed to tackle the challenges Bands, RFE, Random Forest by capturing and interpreting complex patterns and relationships in historical price data. Regressor, Multivariate Analysis, Technical features are important for understanding market trends and thus improving the LSTM. accuracy of stock price predictions. In this paper, we calculate key technical indicators such as Simple Moving Average (SMA), Exponential Moving Average (EMA), Relative Strength Index (RSI), Moving Average Convergence Divergence (MACD), Bollinger Bands, and others. We then focus on selecting the most relevant indicators by employing feature selection methods from these to enhance the extraction of meaningful features reflecting underlying market behavior and increase the probability of more precise prediction. Here, Recursive Feature Elimination (RFE) and Random Forest Regressor-based importance ranking methods have been applied for the feature selection task. To get a better forecast of market price, it is important to capture long- term dependencies and patterns over time. Long Short-Term Memory (LSTM) networks are well- suited for modeling and predicting sequential data like stock prices. By leveraging an LSTM model and taking the selected features, we do a multivariate analysis to forecast stock price based on historical data, identifying the trends fairly accurately with some lags here and there.

Keywords

Machine LearningDeep LearningSMAEMARSIMACDBollinger BandsRFERandom Forest RegressorMultivariate AnalysisLSTM
View Full Article

Article Information

Received

6 September 2025

Accepted

17 October 2025

Published

25 October 2025

ISSN

3067-5804

E-ISSN

3067-5812

Article Type

Research Article

Open Access

Yes – Open Access