ECE 8550 Artificial Intelligence

Revised on 12/20/2024

For the definitive and up-to-date course details, please refer to the syllabus on Canvas.

Description

TL; DR: Applied deep learning with PyTorch for beginners.

This introductory course is designed to immerse beginners into the world of deep learning using PyTorch. Participants will gain hands-on experience in developing different types of deep neural networks, including fully connected, convolutional, and recurrent neural networks. Key aspects of PyTorch programming will be covered, providing learners with the skills to implement neural network architectures effectively. The course will also delve into crucial practices for training and tuning deep neural networks to achieve optimal performance. Additionally, it will address contemporary issues in AI, such as security, privacy, fairness, and explainability, ensuring a well-rounded understanding of the ethical and technical considerations in deep learning applications. This course is ideal for those looking to start their journey in the rapidly evolving field of deep learning.

Topics

  • Deep neural networks
    • fully connect neural network
    • convolutional neural network
    • recurrent neural network
  • PyTorch programming
  • Training and tuning deep neural network
  • Deep learning frontiers
  • Trustworthy AI, e.g., security/privacy/fairness/explainability

Prerequisite

Course Materials

No textbooks are required for this course. I recommend the following book for after-class reading:

  • Aston Zhang, Zachary C. Lipton, Mu Li and Alexander J. Smola. (2020). Dive into Deep Learning. Available Online.

Grading

  • Homework
  • Paper reading
  • Final project
  • Attendance

Late Policy

  • On the due date, the cutoff for on-time submission is 11:59 pm (East Time).
  • Late work is discounted 5% per calendar day late.
  • Late submissions are not accepted after seven calendar days past the original due date and are graded as zero immediately.