Biography

I am a Postdoc in the Department of Electrical and Computer Engineering at University of Victoria, BC, Canada. My research interests focus on the optimization problems in machine learning mainly including Federated Learning.

Teaching 2024 Fall

ECE 403/503 Optimization for Machine Learning

Course Description

The course is designed to introduce students to the core problems in machine learning, starting with an exploration of foundational datasets. Students will gain an understanding of common challenges in machine learning and the approaches used to address them. A key objective is to develop proficiency in unconstrained optimization, focusing on concepts such as gradients, Hessians, and optimality conditions, and their applications in machine learning. The course emphasizes practical optimization algorithms like gradient descent and Newton methods, crucial for improving the efficiency and accuracy of machine learning models. Additionally, it covers supervised learning techniques, including least-squares linear regression and logistic regression for both binary and multi-class tasks. Advanced topics like momentum-accelerated optimization, stochastic gradient descent, and Quasi-Newton methods are also included. The course concludes with feature engineering, multi-class classification, regularization, and unsupervised learning techniques like K-Means clustering and PCA.