报告题目： Towards Robust Multimodal Machine Learning for Healthcare
主讲人：殷可经，香港浸会大学研究助理教授
报告时间：2024年11月19日（星期二）10:00-11:30
报告地点：厦门大学翔安校区西部片区1号楼108
报告摘要：
Clinicians naturally make predictions using multimodal data such as structured clinical records, continuous monitoring signals, and medical images. Strategically combining different data modalities has great potential to improve the accuracy of machine learning models in clinical prediction tasks. However, effective and robust multimodal learning for healthcare data is hindered by a few fundamental challenges, including missing data, incomplete modality, and semantic gaps between modalities. This talk will explore recent advancements aimed at overcoming these obstacles. The first part of the talk will focus on machine learning methods that tackles the not-completely-at-random missing data by joint imputation and prediction for binary electronic health records (EHRs). The second part of the talk will discuss innovative multimodal fusion approaches for clinical predictions, including (1) a disentangled representation learning method for EHRs and chest X-ray images with missing modalities, and (2) a diffusion-based generative model for dynamic latent X-ray generation. By tackling these challenges, we can pave the way for more robust multimodal learning for healthcare, ultimately improving patient care.
报告人简介：
Dr. Kejing Yin is currently a Research Assistant Professor in the Department of Computer Science at the Hong Kong Baptist University. His research focuses on machine learning for healthcare and intensive care big data analytics. He completed his graduate studies at South China University of Technology in 2015 and earned his PhD from Hong Kong Baptist University in 2021. In 2019, he was a visiting student at Georgia Institute of Technology in under the mentorship of Prof. Jimeng Sun. Dr. Yin’s research havs been published in leading artificial intelligence and data mining venues, including NeurIPS, AAAI, and IEEE TKDE. In 2021, he received an Honorable Mention for the Hong Kong Young Scientist Awards. As a Research Assistant Professor, he leads one NSFC research grant and one Health and Medical Research Fund project from Hong Kong Health Bureau.
邀请人：计算机科学与技术系 卢杨助理教授
