Date of Award


Document Type


Degree Name

Master of Science in Information Systems and Technology


Information and Decision Sciences

First Reader/Committee Chair



This culminating experience project investigates the effectiveness of convolutional neural networks mixed with long short-term memory (CNN-LSTM) models, and an ensemble method, extreme gradient boosting (XGBoost), in predicting closing stock prices. This quantitative analysis utilizes recent AAPL stock data from the NASDAQ index. The chosen research questions (RQs) are: RQ1. What are the optimal hyperparameters for CNN-LSTM models in stock price forecasting? RQ2. What is the best architecture for CNN-LSTM models in this context? RQ3. How can ensemble techniques like XGBoost effectively enhance the predictions of CNN-LSTM models for stock price forecasting?

The research questions were answered through a thorough quantitative analysis involving data preprocessing, feature engineering, and model evaluation, using various Python scripts designed for this analysis. The findings are: RQ1. reveals that adjusting hyperparameters, such as learning rates and epochs, significantly improves model performance; RQ2. deemed a multi-layered CNN-LSTM structure with attention mechanisms as the most effective for this use case; and RQ3. showed that XGBoost as an ensemble method did not work as planned, indicating a much more complex interplay between ensemble methods and neural network models. The conclusions are: RQ1. adjusting hyperparameters, such as learning rates and epochs, improves the performance of CNN-LSTM models. RQ2. multi-layered CNN-LSTM architectures with attention mechanisms are the most effective architecture for predicting stock prices. RQ3. ensemble methods like XGBoost, when combined with CNN-LSTM models, did not improve prediction accuracy as expected, suggesting a complex interplay between these techniques. Areas for further study include the automation of hyperparameter tuning techniques such as GridSearch and Bayesian optimization, further exploration of the integration of ensemble methods with neural network models, and the application of CNN-LSTM architectures to other forms of financial data beyond closing stock prices.