Chenliang Xu

pic

Fall 2021 - CSC 298/578: Deep Learning

Basic Information

Meet Time/Location: T/R 09:40-10:55 in Hylan 201

Instructor: Prof. Chenliang Xu
Instructor's Office Hours: Wednesday 11:00-noon on Zoom (may leave early if nobody shows up)

TAs and Office Hours:

Information Flow: We will use Blackboard to make announcements, host lecture notes & assignments, and submit student work. The course ID is CSC298.1.FALL2021ASE. We will use the Discussion Board of Blackboard for class-related discussion and communication. In this way, you can get help fast and efficiently from classmates, TAs, and myself. Please email the TAs or me if it is a personal event. Finally, this course website contains an outline of the course schedule and serves as a syllabus.

Credit Hours: This is a four-credit-hour course. The credit hours are met by weekly lectures (150 minutes), office hours, homework assignments, and the final course project.


COVID-19 Safety and Preventions

Code of Conduct: Check here for classroom "etiquette" expectations. In light of COVID, please:

Everyone should comply with the COVID-19 Community Commitment. In the event of showing symptoms, under quarantine due to exposure, or testing positive for COVID-19, please email me for class accomodations.


Course Description

Much of the recent advances in machine learning and artificial intelligence have been dominated by neural network approaches broadly described as deep learning. This course provides an overview of the most important deep learning techniques covering both theoretical foundations and practical applications. The applications focus on problems in image understanding and language modeling utilizing state-of-the-art deep learning libraries and tools, which will be introduced in the course.

Course Objective: The course aims to provide students with an introduction to deep learning and an in-depth look at current research topics in this field. The course is designed as an upper-level elective course mainly for junior and senior students in Computer Science and Data Science.

Prerequisites: MTH 164 (multidimensional calculus), MTH 165 (linear algebra), and CSC 172 (data structures and algorithms). Programming in Python. Recommended but not required: CSC 246 and CSC 249.

Hardware/Software Requirements: Course projects may require you to have adequate GPU computing resources. This can usually be resolved by:

  1. Using our provided CSUG accounts to access shared GPU resources in CS instructional network; or
  2. Owning an NVIDIA GPU system with 6GB or greater GPU memory; or
  3. Renting a cloud GPU via cloud service providers, e.g., Google Cloud.

In any cases, please test your code in CS instructional network before submission as this is the environment we develop and grade your assignments. Follow this link to request an account if you don't already have one from your other CS courses.


Schedule

All in-person lectures will be automatically recorded by UIT. Recording will start at the Registrar’s designated time for start of class and end +5minutes at the end.

This is a tentative schedule of the course and is subject to modifications over the term.
Part I: Neural Network Basics, Part II: Intermediate Topics, Part III: Advanced Topics.

Dates Tuesday (T) Thursday (R) Notes
8/24,26 No Class: Before Term Introduction
8/31,9/2 Machine Learning Basics I Machine Learning Basics II
9/7,9 Feedforward Neural Networks Backpropagation P-I Out
9/14,16 PyTorch Library (TA) Regularization
9/21,23 Optimization Convolutional Neural Networks P-III Teaming
9/28,30 CNN Architectures Recurrent Neural Networks P-II Out
10/5,7 Recursive, Graph Neural Networks Explain and Visualization P-I Due
10/12,14 Fall Break Object Recognition I
10/19,21 Object Recognition II Face Recognition P-III Proposal
10/26,28 Attention Big Models P-II Due
11/2,4 Network Architecture Search Autoencoders
11/9,11 Generative Adversarial Networks Style Transfer
11/16,18 Video and Multimedia I Video and Multimedia II
11/23,25 Adversarial Attacks Thanksgiving
11/30,12/2 Bias and Fiarness Project Presentation I P-III Due
12/7,9 Project Presentation II No Class: Term Finished

Course Work and Evaluation

Late Work Policy: Deduct 1% total score of the late assignment per hour passing the deadline. E.g., you submit Project I 12 hours passing the deadline, then the max score you will get is 88% of Project I. We will not accept your submission after 72 hours over the deadline. All assignments should be submitted on Blackboard.

Regrading: Any questions about the grading of a piece of work must be raised within one week of the date that the work was returned. In other words, if you do not pick up your work in a timely fashion, you may forfeit your right to question the grading of your work.

Policy on Incomplete: Incomplete if the student missed Project III or any two of the projects.


Textbook and Other Resources

I do not enforce a textbook. For beginners, I recommend reading the book below for a quick start.

For a deep dive into deep learning, I recommend reading:

There are tons of online materials and courses for learning deep learning. This Github repository awesome-deep-learning is an excellent place to get started. You may also learn from deep learning library tutorials and manuals. However, technologies come and technologies go, but insights stay. Therefore, it is more important to understand the principles behind, and you can quickly master any libraries later if needed.


Other Policies

Academic Honesty: All assignments and activities associated with this course must be performed in accordance with the University of Rochester's Academic Honesty Policy. More information is available here. You are expected to read, understand, and follow the policy. Additionally:

Posting homework and project solutions to public repositories on sites like GitHub is a violation of the College’s Academic Honesty Policy, Section V.B.2 “Giving Unauthorized Aid.”

Disabilities: Please see me about your required accommodations as early as possible in the term. The University of Rochester respects and welcomes students of all backgrounds and abilities. In the event you encounter any barrier(s) to full participation in this course due to the impact of a disability, please contact the Office of Disability Resources. The access coordinators in the Office of Disability Resources can meet with you to discuss the barriers you are experiencing and explain the eligibility process for establishing academic accommodations. You can reach the Office of Disability Resources at: disability@rochester.edu; (585) 276-5075; Taylor Hall; www.rochester.edu/college/disability.