Forecasting Ethereum’s Price using ML and DL by Integrating Hybrid Sentiments in Multi-Source Market Data: Leveraging XAI

Satish, Naresh Kumar (2024) Forecasting Ethereum’s Price using ML and DL by Integrating Hybrid Sentiments in Multi-Source Market Data: Leveraging XAI. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (2MB) | Preview
Preview	PDF (Configuration Manual) Download (4MB) | Preview

Abstract

The cryptocurrency market, which is considered to be one of the most volatile markets due to its inconsistencies in the pricing factors, is yet widely used by a larger population incurring losses in most of the cases. To act as a risk assessment factor among the investors, users and other groups, the research study leverages the concept of forecasting Ethereum’s prices by analyzing its social media sentiments like global news headlines, Reddit discussion forums and enhancing the data with hybrid sentimental features derived from VADER, BERT, Text Blob and correlating them with the financial parameters of the Ethereum to build a strong relationship among them and train machine learning models. The study has showcased prediction results of Ethereum using Random Forest, Extreme Gradient Boosting and Long Short-Term Memory models, critically evaluated for various factors and visualized that Extreme Gradient Boosting outperforms the other two models in capturing the complex relationship in the data and presenting a R-squared value of 0.982115. The study has presented the critical evaluation of the models, justification of the model’s results and limitations. To enhance the risk assessment application of the study, the concept of explainable AI has been utilized to have transparency and accountability in the model’s results. Shapley Additive Explanations (SHAP) is incorporated in the research study to explain the XG Boost’s model’s interaction on the features enhancing the reliability of the model. And concluding with some of the limitations of the model’s performance regarding the nature of data.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Simiscuka, Anderson UNSPECIFIED
Subjects:	H Social Sciences > HG Finance Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science H Social Sciences > HG Finance > Money > Digital currency Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions:	School of Computing > Master of Science in Artificial Intelligence
Depositing User:	Ciara O'Brien
Date Deposited:	20 Jun 2025 10:18
Last Modified:	20 Jun 2025 10:18
URI:	https://norma.ncirl.ie/id/eprint/7965

Actions (login required)

View Item