Year

2026

Season

Spring

Paper Type

Master's Thesis

College

College of Arts and Sciences

Degree Name

Master of Science in Mathematical Sciences (MS)

Department

Mathematics & Statistics

NACO controlled Corporate Body

University of North Florida. Department of Mathematics and Statistics

Committee Chairperson

Dr. Fei Heng

Second Advisor

Dr. Elena Buziaianu

Rights Statement

http://rightsstatements.org/vocab/InC/1.0/

Third Advisor

Dr. Ping Sa

Department Chair

Dr. Richard F. Patterson

College Dean

Kaveri Subrahmanyam

Abstract

Liver cirrhosis is associated with substantial morbidity and mortality, making one-year mortality prediction a clinically relevant problem. Using a liver cirrhosis dataset as the motivating application, this thesis evaluates five machine learning classifiers—Logistic Regression, Random Forest, XGBoost, LightGBM, and CatBoost—under five class-imbalance handling strategies: Baseline learning, Random Oversampling, SMOTE-NC, ADASYN, and Cost-Sensitive Learning. Hyperparameter tuning was conducted using randomized search, and predictive performance was assessed over 200 iterations of Monte Carlo Cross-Validation using Accuracy, Precision, Recall, Fl-score, and ROC-AUC.

The results suggest that imbalance-handling strategies can materially affect predictive performance, particularly recall. Because the outcome of interest is death within one year, recall is especially important, as low recall implies failure to identify true high-risk patients. Across the models, baseline learning often achieved high accuracy and, in many cases, high precision, but it generally yielded the weakest recall. In contrast, ADASYN, SMOTE-NC, and Cost-Sensitive Learning generally improved recall, while SMOTE-NC and Cost-Sensitive Learning often produced stronger Fl-scores, indicating a better balance between precision and recall. Overall, the findings suggest that no single imbalance-handling strategy is uniformly optimal across classifiers, but several such strategies can provide a more favorable balance among clinically relevant performance measures in this setting.

Share

COinS
 

Accessibility Statement

This item was created or digitized before April 24, 2027, or is a reproduction of legacy material created before that date. It is preserved in its original, unmodified state specifically for research, reference, or historical recordkeeping. In accordance with the ADA Title II Final Rule, the Library provides accessible versions of archival materials by request. If you are experiencing difficulty accessing the information on the site due to a disability, please submit a request through the following form for assistance.