Spring 2025: Machine Learning [ CS688 ]


General Description and Preliminary List of Topics
Machine learning studies computer algorithms for learning to do things. For example, we might be interested in learning to complete a task, or to make accurate predictions, or to navigate in an unexplored environment. The learning that is being done is always based on some sort of observations or data, such as examples (the most common case in this course), direct experience, or instruction. So in general, machine learning is about learning to do better in the future based on what was experienced in the past. The emphasis of machine learning is on automatic methods. In other words, the goal is to devise learning algorithms that do the learning automatically without human intervention or assistance.

The machine learning paradigm can be viewed as “programming by example.” Often we have a specific task in mind, such as recognizing handwritten digits on an envelope to perform automated mail dispatching. But rather than program the computer with rules to solve the task directly, in machine learning, we seek methods by which the computer will come up with its own program based on examples that we provide.

The course covers key algorithms and theory at the core of machine learning. Particular emphasis will be given to the statistical learning aspects of the field, the mathematical formulations of common tasks, and the techniques used to solve them. We will explore the reasoning behind the formulation of learning algorithms. This will enable students to gain a deeper understanding of the objectives, underlying assumptions, advantages, and limitations of the different techniques.

This is a theory class! Students enrolled in this class must be aware that machine learning is a computer science field that sits at the intersection of different disciplines and it leverages a variety of tools and techniques from optimization theory, probability and statistics, and multivariate linear algebra and calculus. So expect to see a fair amount of material drawn from these disciplines!

Topics include: decision theory, Bayesian theory, curse of dimensionality, dimensionality reduction techniques, classification, clustering, neural networks, kernel methods, mixture models and EM, deep learning. If time permits, additional topics may be covered.

Course Format
Lectures by the instructor. Besides material from the textbook, topics not discussed in the book may also be covered. Research papers and handouts of material not covered in the book will be made available.

Classroom Specifics
I expect students to attend the class. I supplement the textbooks with extensive discussions and material. Students' active participation is very important to succed in this course. Audio recording of the lectures is not allowed.

Communication
We use Piazza for communication, and to enable questions and students' discussions. Class material, handouts, and grades are posted on Blackboard. We use GradeScope for submission of homework assignments.

Grading Policy
Grading will be based on quizzes, homework assignments, and two exams. No make-up quizzes or make-up exams will be offered. The lowest grade on the quizzes will be dropped. Some of the assignments will require programming. No late assignments will be accepted. Graded work must be done on an individual basis, unless otherwise stated by the instructor. No extra work will be offered during the semester or at the end-of-semester for extra credit. If you're not happy with the trajectory of your grade, change something and change it fast! Any deviation from this policy will be considered a violation of the GMU Honor Code.

Grade Breakdown
Quizzes: 20%
Homework assignments: 20%
Midterm: 30%
Final: 30%
Grading Schema
(the threshold may be adjusted for certain grades to take into account the performance of the class as a whole)

Letter Grade Score Range
A+ >=98
A [95, 98)
A- [90, 95)
B+ [85, 90)
B [80, 85)
B- [75, 80)
C+ [70, 75)
C [65, 70)
C- [60, 65)
F <60
Honor Code
This class enforces the GMU Honor Code and the more specific honor code policy special to the Department of Computer Science. You will be expected to adhere to this code and policy.

Disabilities
If you have a documented learning disability or other condition which may affect academic performance, make sure this documentation is on file with the Office of Disability Services and talk to me about accommodations.