Academics

  /  

Courses

  /  

Descriptions

COMP_SCI 496: Mathematical Foundations of Machine Learning

VIEW ALL COURSE TIMES AND SESSIONS

Prerequisites

PhDs students only

Description

This class will meet for 3 hours per week and an additional (and optional) tutorial/review sessions may be scheduled as needed.

• Cross-listed with STAT 435

COURSE DESCRIPTION:
In this course, students are expected to explore some mathematical foundations of modern machine learning under a problem-solving framework. Topics include probability theory, frequentist statistics, Bayesian statistics, tensor algebra, vector calculus, convex and stochastic optimization, sequential optimization, and dynamic programming. This class strongly emphasizes on developing problem-solving skills.

SYLLABUS:
- Motivation and a unified framework for machine learning
- Probability theory, statistical models, statistical inference
- Frequentist and Bayesian Statistics
- Tensor algebra, vector calculus and auto-differentiation
- Convex and stochastic optimization
- Sequential optimization and dynamic programming

EVALUATION:
Students will be evaluated based on in-class problem solving and take-home assignments.

REFERENCE TEXTBOOKS: N/A
REQUIRED TEXTBOOK: No required textbook. In-class lecture notes will be provided.

COURSE COORDINATORS: Prof. Han Liu

COURSE INSTRUCTOR : Prof. Han Liu