Leveraging Pre-trained Models for Machine Learning on Cancer Datasets

Principal Investigator

Name
Daeyoung Hong

Degrees
Ph.D.

Institution
Myongji University

Position Title
Co-Principal Investigator

Email
dyhong@mju.ac.kr

About this CDAS Project

Study
PLCO (Learn more about this study)

Project ID
PLCO-1902

Initial CDAS Request Approval
Apr 30, 2025

Title
Leveraging Pre-trained Models for Machine Learning on Cancer Datasets

Summary
This project aims to investigate the effectiveness of pre-trained machine learning models when applied to cancer datasets. In medical domains annotated data is often limited, making transfer learning a valuable strategy. By utilizing pre-trained models, we aim to evaluate performance improvements in cancer classification tasks.

Aims

- We wish to demonstrate the benefits and limitations of using pre-trained models when applied to cancer datasets.
- We wish to assess fine-tuning strategies to cancer datasets.
- We wish to evaluate how representations learned from general data can be adapted to capture patterns in cancer diagnosis.

Collaborators

Woohwan Jung: Hanyang University