Fall 2023
This seminar course will discuss cutting-edge machine learning (ML) methods in addressing critical classification problems. Specifically, we will cover two types of applications, cybersecurity (e.g., detecting stealthy intrusions and threats) and digital health (e.g., early detection of diseases and treatment predictions). Our discussion will be technical and quantitative, focusing on examining the prediction accuracy, methodology design for evaluation, deployability of the ML solutions, and open research problems.
The format of the course will include lectures, student paper presentations, and semester-long projects. Through discussing state-of-the-arts research papers and completing projects, students will learn to identify pain points in critical classification problems, understand the constraints and common pitfalls in applying machine learning techniques, and be able to design scientifically sound and objective experiments to evaluate advanced machine learning methods.
Suggested prerequisite: CS 5804 (Intro Artificial Intelligence) or CS 5814 ( Introduction to Deep Learning) or equivalent machine learning and data science courses.