Abstract—Stock market prediction is a key objective in financial engineering, requiring advanced analytical methods to model complex market behavior. As global markets grow more dynamic, Machine Learning (ML) and Deep Learning (DL) methods increasingly outperform traditional statistical approaches. This study introduces a structured taxonomy of forecasting techniques and evaluates major regression-based models, including linear regression, logistic regression, and neural architectures such as Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), and hybrid Convolutional Neural Network with Long Short-Term Memory (CNN-LSTM) networks. Five predictive models—Linear Regression (LR), Logistic Regression, RNN, LSTM, and CNN-LSTM—were implemented. LR achieved the strongest performance, with Root Mean Square Error (RMSE) values of 0.334 (training) and 0.304 (testing), while Logistic Regression performed the worst with RMSE values of 0.463 (training) and 0.487 (testing). RNN and LSTM produced higher errors than LR (RMSE 0.355 and 0.383), showing that increased model complexity does not guarantee higher predictive accuracy. The approach applies advanced preprocessing, including Z-score normalization and temporal sequence structuring. Results indicate that predictive performance depends heavily on data characteristics and market conditions. While previous studies exist, this work offers novelty by applying a structured ML-DL taxonomy to an underexplored emerging market (Saudi Aramco/Tadawul), implementing a reproducible preprocessing framework, and demonstrating that simpler models can outperform more complex architectures in markets with limited volatility and data depth.
 
Keywords—deep learning, machine learning, prediction, Artificial Intelligence (AI) methods, stock market, regression, taxonomy


Copyright © 2026 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (CC BY 4.0).