• Version
• Download 14
• File Size 595.57 KB
• File Count 1
• Create Date 31 January 2026
• Last Updated 31 January 2026

Stock Prediction using Machine Learning

Dr. N.Prabakaran
Department of Electronics and Communication Engineering Koneru Lakshmaih Educational Foundation
Vaddeswaram, Guntur, India

B.Tanishk
Department of Electronics and Communication Engineering Koneru Lakshmaiah Educational Foundation
Vaddeswaram, Guntur, India 2100040052ece@gmail.com

D.Bhaskar
Department of Electronics and Communication Engineering Koneru Lakshmaiah Educational Foundation
Vaddeswaram, Guntur, India 2100040295@kluniversity.in

ABSTRACT: stock prediction is an important process for providing a business with insight into strategies. Deep learning models such as
Convolutional Neural Network (CNN), Recurrent Neural Networks (RNN), and Artificial Neural Networks (ANN) are currently very prevalent in predictive analytics. However, depending on the structured nature of data, much simpler models such as Multiple Linear Regression (MLR) can be more successful. In this paper we assess and compare the application MLR works versus deep learning for
forecasting sales with financial and demographic data. Comprehensively we find MLR outperforms all three deep learning model alternatives across multiple evaluation criteria, and highlight how model choice should be based on the characteristics of the data over
model complexity.

Keywords: Sales forecasting, deep learning, multiple linear regression, CNN, RNN, ANN, machine learning, predictive modelling.