SEMIPARAMETRIC METHODS IN DATA SCIENCE
Review of statistical convergence modes, M-estimation, and basics of Hilbert space. Introduction of how to derive the nuisance tangent space, its complement, and the corresponding efficient influence function, from the geometric perspective of semiparametric models. Introduction of how to estimate nuisance functions using machine learning methods, and their implementations in R and/or Python. Introduction of a variety of semiparametric models in missing data analysis, causal inference, dimension reduction, precision medicine, semi-supervised learning, transfer learning and domain adaptation.
Prerequisites & course information
Requisites: Graduate/professional standing Course Designation: Grad 50% - Counts toward 50% graduate coursework requirement Repeatable for Credit: No Last Taught: Fall 2024 Learning Outcomes: 1. Derive the nuisance tangent space, its complement, and the corresponding efficient influence function in semiparametric models Audience: Graduate 2. Apply a variety of semiparametric methods and models in applications ranging from biomedical studies to social sciences Audience: Graduate 3. Perform machine learning algorithms for estimating nuisance functions in software such as R and/or Python Audience: Graduate Learning Outcomes: 1. Derive the nuisance tangent space, its complement, and the corresponding efficient influence function in semiparametric models Audience: Graduate 2. Apply a variety of semiparametric methods and models in applications ranging from biomedical studies to social sciences Audience: Graduate 3. Perform machine learning algorithms for estimating nuisance functions in software such as R and/or Python Audience: Graduate
Sources
- Course catalogue ↗2026-2027
- Catalogue entry ↗
Catalogue variants
- SEMIPARAMETRIC METHODS IN DATA SCIENCE · 3 credits.