Applications of Machine Learning in Medicine
Program

Transform healthcare with machine learning.

Applications of Machine Learning in Medicine is a two-course professional program for clinicians, researchers, and health data professionals who want to apply machine learning to real healthcare challenges.

This online, self-paced program covers the essentials of working with healthcare databases, knowledge graphs, and structured and unstructured data, as well as techniques for electronic phenotyping and time series analysis.

After completing this program, you will be able to:

• Evaluate healthcare databases and the U.S. healthcare data ecosystem, identifying data provenance, sources of error and bias, and strategies to assess and improve data usability.
• Build structured clinical datasets from longitudinal timelines by defining units of observation, engineering features, handling missing data, and mitigating data cascades.
• Process unstructured clinical text using natural language processing (NLP) and de-identification techniques, and apply knowledge graphs such as the Unified Medical Language System (UMLS) for concept mapping in healthcare data.
• Analyze healthcare time-series data, accounting for non-stationarity and the timing of exposures and outcomes to model patient trajectories.
• Design and validate electronic phenotypes and reproducible cohorts using rule-based and machine learning approaches, supported by chart review.
• Develop, tune, and evaluate supervised learning models, including risk scores and predictive models, using appropriate loss functions, bias–variance tradeoffs, and cross-validation.
• Apply deep learning methods to electronic health records (EHR) and medical data, including Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), attention mechanisms, Transformers, and transfer learning.
• Plan and execute clinical AI deployment by defining use cases, understanding steps beyond prediction, ensuring interpretability and generalizability, and addressing ethics and liability.