Jehyun Park

ML/Deep Learning Researcher
CS Student
Sungkyunkwan University

Research Interests

My research focuses on several key areas in machine learning and optimization:

Optimization & Generalization

  • Convex optimization techniques for model training
  • Resolving generalization problems through mathematical approaches
  • Exploration of the “grokking” phenomenon in neural networks
  • Application of Taylor series, Fourier and Laplace transformations

Neural Fields & Learning Mechanisms

  • Understanding memorization patterns in deep learning
  • Mechanisms behind neural network learning processes
  • Implementation and optimization of Neural Fields
  • Balancing memorization vs. generalization in models

Machine Learning for Industrial & Medical Applications

  • Developing models for continuous and tabular data domains
  • Feature engineering techniques for medical datasets
  • Handling of imbalanced data in healthcare applications
  • Vector representation methods for categorical medical features

Current Research

AI Research at The Catholic University of Korea St. Vincent’s Hospital

Sep 2024 — Feb 2025

My ongoing research involves developing machine learning algorithms for medical applications:

  • Feature engineering with medical data
  • Implementation of K-NN imputers for missing data
  • Dimensionality reduction with t-SNE for visualization
  • Imbalanced learning techniques with specialized XGBoost variants

Recommendation Systems Using LLMs

Jul 2024 — Jan 2025

Working with LLM-based recommendation systems in healthcare:

  • Transfer learning with large language models
  • Adaptation of LLMs to medical domains
  • Integration of Graph Attention Networks (GAT)
  • Implementation of Chain-of-Thought (COT) reasoning

Publications

Coming soon

Research Experience

Undergraduate Research at iisLab SKKU

Jun 2024 — Aug 2024

  • Participation in Computer Vision research seminars
  • Natural Language Processing studies
  • Presentations on unsupervised learning techniques
  • Research on optimization methods and RNNs

Artificial Intelligence College Conference (TNT SKKU)

Mar 2024 — Jun 2024

  • Weekly study groups on AI advancements
  • Presentations on self-RAG systems
  • Analysis of Mistral attention mechanisms
  • Foundational studies in Reinforcement Learning

Industry-Academia Collaboration (SOYNET)

Jun 2023 — Aug 2023

  • AI model optimization techniques
  • Inference acceleration for real-time applications
  • Implementation of CUDA optimizations
  • Person Re-ID model enhancement

Collaborations

I am open to research collaborations in the following areas:

  • Machine learning for healthcare applications
  • Optimization techniques for deep learning
  • Neural networks for financial modeling
  • Recommendation systems with LLMs

If you’re interested in potential collaboration, please contact me.