时间：2025年1月3日下午4点
地点：新校区信息楼416会议室
报告人: 龙琦
个人简介：
龙琦，现任美国宾夕法尼亚大学医学院生物统计学教授，工程学院计算机学教授，沃顿商学院统计学教授。同时兼任生物统计，流行病，及信息学系系副主任；癌症数据科学中心创始主任；癌症研究院主管数据科学和人工智能的副院长；生物医学信息研究院副院长。 当选美国科学促进会会士美国统计学会会士, 国际统计学会会士，以及美国医学信息学会会士。长期从事与精准医疗和健康有关的统计学, 信息学, 机器学习, 和人工智能的研究, 包括医疗大数据分析和推断(多组学数据, 电子病历数据, 影像数据等), 缺损数据, 因果分析, 贝叶斯方法, 临床实验，数据隐私，算法公平，以及大语言模型等方向。发表的杂志包括国际知名期刊Nature Medicine, Nature Communications, JAMA Oncology, Cancer Research, PNAS, 统计学顶刊Annals of Statistics, Journal of the American Statistical Association, 以及AI/ML顶会 ICML, NeurIPS。
报告摘要：
Rich electronic health records (EHR) data offer remarkable opportunities in advancing precision medicine, they also present daunting analytical challenges. Multi-modal data in EHR that are recorded at irregular time intervals with varying frequencies include structured data such as labs and vitals, codified data such as diagnosis and procedure codes, and unstructured data such as clinical notes and pathology reports. They are typically incomplete and fraught with other data errors and biases. What’s more, data gaps and errors in EHRs are often unequally distributed across patient groups: People with less access to care, often people of color or with lower socioeconomic status, tend to have more incomplete data in EHRs in United States. Such data bias, if not adequately addressed, would lead to biased results and exacerbate health inequities. In this talk, I will share my research group’s recent works on responsible AI/ML models including large language models (LLMs) for addressing these challenges. Since LLMs are themselves plagued by various biases, I will also briefly discuss our ongoing research on developing rigorous statistical and ML methods for mitigating pitfalls and risks of LLMs.
