Current Academic Work
Ongoing research, coursework, and academic projects as part of my Master's degree in Data Science.
Master's Thesis Research
Advanced Machine Learning Techniques for Predictive Analytics in Complex Healthcare Datasets
My thesis research focuses on developing novel machine learning approaches to improve predictive accuracy in healthcare analytics. The work combines deep learning architectures with traditional statistical methods to create robust models that can handle the complexity and heterogeneity of medical data.
Research Objectives:
- • Develop hybrid ML models combining neural networks with statistical approaches
- • Improve prediction accuracy for patient outcome forecasting
- • Address data quality and missing value challenges in healthcare datasets
- • Create interpretable models for clinical decision support
- • Validate approaches across multiple healthcare institutions
Methodology:
The research employs a multi-phase approach combining literature review, algorithm development, experimental validation, and real-world testing. I'm working with anonymized patient data from three major healthcare systems to ensure generalizability of results.
Thesis Details
Dr. Sarah Johnson, PhD
Dr. Michael Chen, MD, PhD
Spring 2025
Data Collection Complete
Progress
Current Coursework (Fall 2024)
Advanced Deep Learning (CS 7643)
Exploring cutting-edge deep learning architectures including Transformers, GANs, and reinforcement learning. Focus on both theoretical foundations and practical implementations.
Key Topics:
- • Attention mechanisms and Transformer architectures
- • Generative Adversarial Networks
- • Deep Reinforcement Learning
- • Neural Architecture Search
Statistical Learning Theory (STAT 8803)
Mathematical foundations of machine learning, covering PAC learning, VC theory, and generalization bounds. Emphasis on theoretical understanding of learning algorithms.
Key Topics:
- • PAC learning framework
- • VC dimension and Rademacher complexity
- • Generalization bounds
- • Online learning algorithms
Research Methods in Data Science (CS 8001)
Comprehensive course on research methodology, experimental design, and scientific writing for data science research. Includes ethics and reproducibility considerations.
Key Topics:
- • Experimental design and hypothesis testing
- • Research ethics and bias mitigation
- • Reproducible research practices
- • Scientific communication
Advanced Data Visualization (CS 7450)
Theory and practice of data visualization, covering perception, interaction design, and advanced visualization techniques for complex datasets.
Key Topics:
- • Visual perception and cognition
- • Interactive visualization design
- • High-dimensional data visualization
- • Visualization evaluation methods
Current Research Projects
Federated Learning for Healthcare Data Privacy
Collaborative research project exploring federated learning approaches to train machine learning models on distributed healthcare datasets while preserving patient privacy. Working with industry partners and multiple healthcare institutions.
Collaborators:
Dr. Sarah Johnson (Advisor), Dr. Michael Chen (Co-Advisor), Healthcare Analytics Lab, TechCorp Research Division
Status
Active Research
Expected completion: Spring 2025
Interpretable AI for Financial Risk Assessment
Developing explainable machine learning models for financial risk assessment that provide both high accuracy and interpretable insights for regulatory compliance and decision-making.
Collaborators:
Dr. David Park, Financial Analytics Research Group, FinTech Innovation Lab
Status
Paper Submitted
Under review at ICML 2025
Academic Activities & Service
Teaching & Mentoring
Graduate Teaching Assistant
Introduction to Data Science (CS 4803)
Fall 2024
Undergraduate Research Mentor
Supervising 3 undergraduate students on ML projects
2024 - Present
Professional Service
Reviewer
ICML 2025, NeurIPS 2024 Workshop
2024
Student Representative
Graduate Student Council, Data Science Program
2023 - Present
Academic Resources
Access to my research materials, course notes, and academic collaborations.