INFO 5375: Health Tech Oriented Machine Learning

Jan 23-May 8 – Tuesdays and Thursdays 9:45—11:00 A.M.

Instructor: Fei Wang (few2001@med.cornell.edu)

Course Objective: This course introduces students the various types of health data, including patient clinical records, medical images, physiological and vital signals from wearable sensors, multi-omics, etc. and how to use machine learning algorithms to analyze these data and help with real world health problems such as patient screening, risk modeling, disease subtyping and precision medicine. The technical topics to be introduced in this class include classic supervised and unsupervised learning, network analysis, probabilistic modeling, deep learning, transfer learning, algorithmic fairness and interpretability. We will also invite clinicians or researchers working in the health industry to deliver guest lecturers in the class. The students will gain hands-on experience on analyzing real world health data during course assignments and projects.

Credits: 3 credits.

Teaching Assistants: Emily Tseng (et397@cornell.edu).

Course Format:
  • 75 minutes lectures twice a week (including guest lectures)
  • 2 Group Presentations and 2 Group Debates
  • 1 final project (in groups, 2-5 students per group)

  • Prerequisites: Machine Learning, Python Programming

    Date Content Presenter Materials Assignments
    01/24/2023 Introduction of machine learning Zilong Bai    
    01/26/2023 Introduction of medicine and healthcare Fei Wang    
    01/31/2023 Machine learning in clinical risk prediction Zhenxing Xu   Group formation due
    02/02/2023 Generalizability of machine learning models in clinical risk prediction settings Suraj Rajendran    
    02/07/2023 Multi-modal learning Benjamin Clicksberg    
    02/09/2023 Federated Learning Fei Wang    
    02/14/2023 Do we need model interpretability in clinical risk prediction? Group Debates Paper 1, Paper 2, Paper 3, Paper 4, news Group 1,2 vs. 3,4
    02/16/2023 Algorithmic bias Weishen Pan    
    02/21/2023 Reporting guidelines and regulatory policies Group Presentation Consort AI, MIMINAR, PRISMA AI, GDPR, NAM doc, FDA White Paper Group 5,6,7
    02/23/2023 Clinical natural language processing Group Presentation Paper 1, Paper 2, Paper 3, Paper 4, Paper 5, Paper 6 Group 1,2,3,4
    02/28/2023 Break      
    03/02/2023 Implementation of AI models in clinical workflow Yiye Zhang    
    03/07/2023 Causal inference for biomedicine Jacqueline Maasch    
    03/09/2023 Machine learning for critical care Ed Schenck    
    03/14/2023 Will LLM fundamentally change medicine in a short time? Group debates Paper 1, Paper 2, Paper 3, Paper 4, PubMed GPT, Blog Group 5 LLM will change medicine immediately
    Group 6 LLM will not change medicine immediately, but wit will in the long run
    Group 7 LLM will not change medicine at all
    03/16/2023 Canceled Canceled    
    03/21/2023 Genetics and genomics Manqi Zhou    
    03/23/2023 Machine learning for rheumatology Bella Mehta   Project Proposal Due
    03/28/2023 Trial Emulation Chengxi Zang    
    03/30/2023 AI for pharmacology with RWD Ying Li    
    04/04/2023 Break      
    04/06/2023 Break      
    04/11/2023 Other topics: safety and eligibility design Group Presentation Paper 1, Paper 2, Paper 3, Paper 4, Paper 5, Paper 6 Group 1,3,5
    04/13/2023 Where would machine learning be more likely to succeed in innovating pharma R&D? Molecule design or clinical trials? Group Debate Paper 1, Paper 2, Paper 3, News, Mckinsey Report, Deloitte Report Group 2,4 (molecule design) vs. 6,7 (clinical trials)
    04/18/2023 Behavioral & mental health Dan Adler    
    04/20/2023 Machine Learning in Psychotherapy: Technical Opportunities and Ethical Challenges Emily Tseng    
    04/25/2023 Wearables and Consumer Health - Cardiovascular Health Group Presentations Paper 1, Paper 2, Paper 3, Paper 4, Paper 5, Paper 6 Group 2,4,6,7
    04/27/2023 Which is more important for a consumer health product? Technology, user experience or market need? Group Debates Paper 1, Paper 2, Paper 3, IQVIA report, Consumer Health Laws, Merket Report Group 1 Technology is the most important
    Group 3 User experience is the most important
    Group 5 Market need is the most important
    05/02/2023 Public Health/Epidemiology Yongkang Zhang    
    05/04/2023 Final Project Presentation Students   Final Project Presentation
    05/08/2023 Wrap up Fei Wang   Final Project Report Due

    © Fei Wang