Deep transfer learning for clinical decision-making based on high-throughput data: comprehensive survey with
Muhammad Toseef1, Olutomilayo Olayemi Petinrin1, Fuzhou Wang1
1Department of Computer Science, City University of Hong Kong, Hong Kong SAR.
Briefings in Bioinformatics
|July 16, 2023
Summary
Transfer learning enhances machine learning for biomedical research by applying knowledge from preclinical data to clinical predictions. This review highlights deep transfer learning
Area of Science:
- Biomedical Informatics
- Machine Learning
- Precision Medicine
Background:
- Omics data fuels biomedical research and precision medicine.
- Limited clinical training data hinders machine learning model performance.
- Transfer learning addresses data scarcity by transferring knowledge across domains.
Conclusions:
- Deep transfer learning is a powerful technique for overcoming data limitations in clinical research.
- Leveraging preclinical data via transfer learning can improve the prediction of clinical outcomes and drug responses.
- This review provides a detailed analysis and roadmap for future research in this domain.
