DS5220: Supervised Machine Learning and Learning Theory

Course overview

This course provides an introduction to supervised machine learning and its theoretical underpinning. Topics include classical and modern methods for learning from labeled datasets, including regression and classification problems, and how to think about regularization and model selection. Both linear and non-linear learning methods (e.g., neural nets) will be discussed. This course will also provide an overview of the statistical learning theory behind these methods, whenever applicable. By the end of the course, the student is expected to understand the principles of how to apply supervised learning methods, and be able to solve supervised learning problems.

Course requirement:

  • Linear algebra; Probability; Calculus.


  • Time: Mon/Wed 2:50pm - 4:30pm

  • Location: WVG 106

  • Stuff:

    • Hongyang Zhang. OH: Monday 5pm to 6pm (or by email appointment).

    • Saurabh Parkar.

    • Haotian Ju. OH: Thursday 3pm to 5pm.

    • Alina Sarwar. OH: Friday noon to 2pm.

Grading: Four problem sets (45%), one final project report (25%), final exam (25%), and class participation (5%).