Using Hybrid Machine learning Models for Stock Price Forecasting and Trading - A Thesis Submitted by Ahmed Khalil

Aug 5, 2026

Trading stocks of publicly traded companies in stock markets is a demanding problem because investors aim to maximize returns while limiting risks. Research therefore integrates fundamental and technical analysis and increasingly applies machine learning and algorithmic trading, leveraging large data volumes for price forecasting and execution. This study uses four machine learning models—LSTM, W-LSTM, LSTM-ARO, and W-LSTM-ARO—to predict next-day prices, evaluate accuracy with MSE, MAE and R2, and assess profit generation across six NYSE-listed companies.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages40

Reading Time1 h 41 min

AccessFree to download

Chapter one: Introduction

Chapter two: Literature review

Chapter three: Data & Methodology

3.1 Models

3.2 LSTM Hyperparameters

3.3 Data Set

3.4 Evaluation criteria

3.5 Trading strategies

Chapter four: Discussion & Results

Chapter five: Conclusion and future work

References

Appendix

American University in Cairo AUC Knowledge Fountain | Theses and Dissertations | Student Research | | --- | --- | | Spring 5-8-2024
Using Hybrid Machine learning Models for Stock Price Forecasting and Trading.
Ahmed Khalil
[aawkhalil@aucegypt.edu](aawkhalil@aucegypt.edu)
Follow this and additional works at: [https://fount.aucegypt.edu/etds](https://fount.aucegypt.edu/etds)
Part of the Business Intelligence Commons, Management Information Systems Commons, and the Portfolio and Security Analysis Commons | | Recommended Citation APA Citation Khalil, A. (2024) . Using Hybrid Machine learning Models for Stock Price Forecasting and Trading. [Master's Thesis, the American University in Cairo] . AUC Knowledge Fountain. [https://fount.aucegypt.edu/etds/2285](https://fount.aucegypt.edu/etds/2285) MLA Citation Khalil, Ahmed. Using Hybrid Machine learning Models for Stock Price Forecasting and Trading.. 2024. American University in Cairo, Master's Thesis. AUC Knowledge Fountain. [https://fount.aucegypt.edu/etds/2285](https://fount.aucegypt.edu/etds/2285) This Master's Thesis is brought to you for free and open access by the Student Research at AUC Knowledge Fountain. It has been accepted for inclusion in Theses and Dissertations by an authorized administrator of AUC Knowledge Fountain. For more information, please contact [thesisadmin@aucegypt.edu](thesisadmin@aucegypt.edu). Using hybrid Machine learning models for stock price forecasting and trading. A Thesis Submitted by Ahmed Khalil to the Masters of Science in Finance Graduate Program 5th of February 2024 In partial fulfillment of the requirements for the degree of Masters Table of content Chapter one: Introduction 5 Chapter two: Literature review 8 Chapter three: Data & Methodology 12 3.1 Models 12 3.1.1 LSTM 12 3.1.2 W-LSTM 13 3.1.3 LSTM-ARO 14 3.1.4 W-LSTM-ARO 14 3.2 LSTM Hyperparameters 14 3.3 Data Set 15 3.4 Evaluation criteria: 16 3.4.1 Mean Absolute Error (MAE) 16 3.4.2 R-Squared score (R2) 16 3.4.3 Mean Squared Error (MSE) 17 3.5 Trading strategies: 17 3.5.1 Machine learning trading signals 17 3.5.2 Technical Indicator EMA signals 17 3.5.3 Technical Indicator RSI signals 17 Chapter four: Discussion & Results 18 Chapter five: Conclusion and future work 25 References 27 Appendix 30 Fig1: LSTM Model forecasting price against real price. 30 Fig2: W-LSTM Model forecasting price against real price 30 Fig3: LSTM-ARO Model forecasting price against real price 31 Fig4: W-LSTM-ARO Model forecasting price against real price 31 Fig5: Mean Absolute Error (MAE) 32 Fig6: Mean Squared Error (MSE) 32 Fig7: R-Squared (R2) 32 Fig8: Trading revenues/losses generated by machine learning models. 33 Fig9: Trading revenues/losses generated by machine learning models and other trading strategies. 33 Python Libraries used. 34 Calculating EMA & RSI technical indicators 34 Wavelet decomposition code. 34 Preparing data for LSTM Model 35 Early stop configuration 36 ARO model 36 LSTM Model 38 Abstract: Trading stocks of publicly traded companies in stock markets is a challenging topic since investors are researching what tools can be used to maximize their profits while minimizing risks, which encouraged all researchers to research and test different methods to reach such goal. As a result, the use of both fundamental analysis and technical analysis started to evolve to support traders in buying and selling stocks. Recently, the focus increased on using Machine learning models to predict stock prices and algorithmic trading as currently there is a huge amount of data that can be processed and used to forecast and trade stocks. The focus of this paper is to use four machine learning models to forecast next day stock prices and trade stocks accordingly. The models used are LSTM model, W-LSTM model in which Wavelet analysis is used to remove the noise of the time series dat

FAQs

What is the core goal of the thesis?

To use hybrid machine learning models to forecast next-day stock prices and trade accordingly, then evaluate both prediction quality and profit generation.

Which machine learning models are used for forecasting?

LSTM, W-LSTM, LSTM-ARO, and W-LSTM-ARO, where wavelet analysis removes noise and ARO optimization selects hyperparameters.

How are model predictions evaluated and how are trading strategies tested?

Accuracy is measured using MSE, MAE, and R2, and trading performance is assessed by comparing revenues or losses generated by the models against other strategies such as buy-and-hold, RSI, and EMA.

Aug 5, 2026

Upload by Ava Thompson