Graduate level
Fall 2025
Lectures: Saturday 14-16 - Monday 10-12
Instructor: Dr. A. Abdollahpouri
Course Description
This
course provides a rigorous foundation in the principles and practices
of deep learning, moving beyond superficial applications to a
fundamental understanding of the field's theoretical underpinnings. We
will critically examine the architectural principles of modern neural
networks, including convolutional and recurrent networks, attention
mechanisms, and transformer models, exploring both their
representational power and their limitations. The curriculum emphasizes
a mathematical framework, delving into optimization in high-dimensional
non-convex spaces, regularization strategies, and the challenges of
generalization. By integrating theoretical concepts with hands-on
implementation of state-of-the-art models, this course equips students
not merely to use existing tools, but to innovate, critically evaluate
research, and contribute to the advancing frontier of deep learning.
Lecture notes
| Physiological Aspects | Lecture0 |
| Introduction | Lecture1 |
| Lecture2 | |
| Lecture3 | |
| Lecture4 | |
| Lecture5 | |
| Lecture6 |
Grading Policy
Homeworks……20% | |
Project .….. 20% | |
Final Exam…..55% | |
Class Participation.....5% |
Assignments
Homework1 Due date:Homework2 Due date:
Homework3 Due date: