Publication Date
Spring 2026
Degree Type
Master's Project
Degree Name
Master of Science in Computer Science (MSCS)
Department
Computer Science
First Advisor
Maryam Khazaei
Second Advisor
Sayma Akther
Third Advisor
Matthew Morozov
Keywords
Heart Attack, Brain Tumor, Machine Learning, Magnetic Resonance Imaging, Health, Medical Diagnosis
Abstract
Heart attacks and brain tumors are serious medical conditions that can threaten a patient’s life if the diagnosis and treatments are delayed. The goal of this thesis is to evaluate machine learning models to improve the accuracy and speed of the diagnosis, enabling patients to start treatments earlier. For the heart attack risk prediction task, a tabular dataset was used, and the selected models for the task were K-Nearest Neighbors and TabNet. For the Brain Tumor classification, the models Convolutional Neural Network and EfficientNetB0 were applied for analyzing brain magnetic resonance imaging scans. Results show that the TabNet model performed better than K-Nearest Neighbors in heart attack risk prediction. For the Brain Tumor classification task, EfficientNetB0 with transfer learning showed improved performance compared to the standard Convolution Neural Network.
Recommended Citation
Tsai, Rachel, "Performing Data Analysis and Machine Learning Techniques on Health Data" (2026). Master's Projects. 1806.
DOI: https://doi.org/10.31979/etd.macf-vjq8
https://scholarworks.sjsu.edu/etd_projects/1806
