Computing I
BIOSTAT 236
Subject & Catalog Number
Course Information
Description
This is an introductory graduate-level course focusing on statistical computing. It seamlessly blends critical programming and statistical computing concepts, designed to complement the collaborative and reproducible biostatistics research. The curriculum covers a wide array of topics, including (1) optimization techniques such as Gradient Descent (GD), Stochastic Gradient Descent (SGD), Newton's Method, Alternating Direction Method of Multipliers (ADMM), and constrained optimization; (2) sampling methods like Importance Sampling, Markov
Chain Monte Carlo (MCMC), Langevin Dynamics, and stable diffusion alongside foundational linear algebra applications through QR and Cholesky Decomposition, and (3) dynamic programming. Beyond the algorithm aspects, the course emphasizes practical skill development, from coding proficiency and essential best practices to the use of computational tools such as Cluster and Linux environments, Git for version control, and techniques for reproducible research using RMarkdown and Jupyter Notebooks. It also covers software development practices with a focus on developing R packages, visualization and web design to effectively communicate findings, and exploring innovative coding approaches, including automation through advanced tools like ChatGPT. This comprehensive course is designed to equip students with the necessary tools and knowledge for navigating the complexities of statistical computing, with a strong emphasis on practical skills like reproducible research, effective coding practices, and the utilization of modern computational tools.
Course Notes
Offered jointly with the School of Public Health as BST236.
Available for Harvard Cross Registration
