This course addresses the formulation of prediction problems, with a specific focus on empirical risk minimization. It covers the performance or error analysis of statistical and computational learning methods. We will develop the necessary concepts and techniques for theoretical analysis, including consistency, concentration inequalities, uniform convergence, empirical risk minimization, and covering number.

  • ⏲️ Lectures: Fri 10:30AM - 12:15PM
  • 🎒 Lecture/Recitation Location: Lady Shaw Bldg G21

Homepage GitHub

All students welcome: We are happy to have auditors in our lectures.

📋 Reference Textbooks

The following textbooks serve as relevant resources, though none perfectly align with the scope and content of our course.

  • Koltchinskii, V. (2011). Oracle inequalities in empirical risk minimization and sparse recovery problems. Springer.
  • Vaart, A. V. D., and Wellner, J. A. (1997). Weak convergence and empirical processes with applications to statistics. Springer.