24WI - Developing and Deploying Transparent and Reproducible Algorithms for Public Health

Public health predictive algorithms and models are becoming increasingly complex which poses a challenge for reproducibility, transparency and use in practice (deployment). Inefficiencies, biases and errors will occur if algorithm developers cannot report their algorithms clearly, in both human and machine-readable formats. This course introduces the concept of “data pipelines”, software libraries and standards for generating algorithms that can be easily used by others in validation studies and application. Trainees will have the opportunity in the hands-on component to create an algorithm and deploy it as a web application (web API).

Learners are required to have proficiency in R for this course.