CS782
Topics in ML: Deep Learning Generative Models (Spring 2023)

Meeting Time and Location:
T 4:30 pm - 7:15 pm , Online (Blackboard Collaborate)

Instructor: Prof. Daniel Barbará.
Email: dbarbara (at) gmu (dot) edu
Office: Eng. Bldg 4420
Office hour: by appointment

Graduate TA: TBA

Course Home Page
CS782

Overview
This course covers Generative Models in Deep Learning

Generative models are one of the most exciting and fast-growing fields in Machine Learning Wired article

In this class we cover their fundamentals, the challenges of training them, and their applications. Topics include:

Generative Modelling vs. Discriminative Modeling
Brief Introduction to Deep Learning
GANs: Adversarial models and principles.
Instabilities of GAN training.
Training methods (WGAN, Wasserstein-GAN, others).
Varieties of GANs (CGAN, DCGAN, etc.).
GAN applications: cycle-GAN,
Variational Autoencoders.
Applications.

Prerequisites: CS 688.

No textbook: supporting technical articles

Class Attendance
Required. Please arrive on time. I expect to start at 4:30pm sharp; Please participate in class! Ask questions if there is something you don't understand.

Grading Policies

End-of-semester numeric scores will be weighted as follows (tentative plan):

50% Assignments
10% paper presentation
30% Final project
10% Participation (measured by submission of activities reports)
Late Policy
No late assignments will be accepted.
No smartphones, LAPTOPS, TABLETS, or recorders allowed in class. Lectures cannot be recorded without special permission
Honor Code
You are expected to abide by the honor code. Please refer to GMU Honor Code. Any violation of the honor code will result in a zero of the assignment/exam, and may result in an F for the class.
School Calendar

Disability Services

CS782 Topics in ML: Deep Learning Generative Models (Spring 2023)

CS782
Topics in ML: Deep Learning Generative Models (Spring 2023)