CS 4803 / 7643 Deep Learning
Fall 2018, TR 4:30  5:45 pm, CCB 16
Course Information
This is an exciting time to be studying (Deep) Machine Learning, or Representation Learning, or for lack of a better term, simply Deep Learning!
Deep Learning is rapidly emerging as one of the most successful and widely applicable set of techniques across a range of domains (vision, language, speech, reasoning, robotics, AI in general), leading to some pretty significant commercial success and exciting new directions that may previously have seemed out of reach.
This course will introduce students to the basics of Neural Networks (NNs) and expose them to some cuttingedge research. It is structured in modules (background, Convolutional NNs, Recurrent NNs, Deep Reinforcement Learning, Deep Structured Prediction). Modules will be presented via instructor lectures and reinforced with homeworks that teach theoretical and practical aspects. The course will also include a project which will allow students to explore an area of Deep Learning that interests them in more depth.
 Instructor
 Dhruv Batra
 Teaching Assistants
 Nirbhay Modhe Office Hours Wednesdays 5pm6pm CCB 2nd floor
Erik Wijmans Office Hours Tuesdays 12pm1pm CCB 2nd floor
Harsh Agrawal Office Hours Mondays 3pm4pm CCB 2nd floor
Michael Cogswell Office Hours Mondays 5pm6pm CCB 2nd floor  Class meets
 Tuesday, Thursday 4:30  5:45 pm, College of Computing Building Classroom 16
 Piazza
 piazza.com/gatech/fall2018/cs48037643
 Canvas
 gatech.instructure.com/courses/28059
 Gradescope
 gradescope.com/courses/22096
 Staff Mailing List
 cs48037643f18staff@googlegroups.com
Schedule
 Methods for Interpreting and Understanding Deep Neural Networks
 Network Dissection: Quantifying Interpretability of Deep Visual Representations
Date  Topic  Optional Reading  
W1: Aug 21 
Class Administrativia. HW0 out Slides (pptx), Slides (pdf). 

W1: Aug 23 
Image Classification and kNN. Slides, Slides (annotated). Supervised Learning notes, kNN notes. 

W2: Aug 28 
Linear Classifiers, Loss Functions. Slides, Slides (annotated). 

W2: Aug 30 
Regularization, Neural Networks. Slides, Slides (annotated). 

W3: Sep 4 
Optimization, Computational Flow Graphs, and Backprop. Slides, Slides (annotated), Gradients notes. 

W3: Sep 6 
Guest Lecture by Peter Anderson: Training Neural Networks 1. Slides. HW0 due 09/05 11:55pm. 

W4: Sep 11 
HW1 out No class. ECCV. 

W4: Sep 13  No class. ECCV.  
W5: Sep 18 
Forward mode vs Reverse mode Autodiff. Slides, Slides (annotated). 

W5: Sep 20 
Convolutional Neural Networks (CNNs). Slides, Slides (annotated), Gradients notes. 

W6: Sep 25 
Convolutional Neural Networks (CNNs) Part 2. Slides, Slides (annotated). 

W6: Sep 27 
Guest Lecture by Peter Anderson:
Training Neural Networks 2. Slides 

W7: Oct 2 
Convolutional Neural Networks (CNNs) Part 3. Slides, Slides (annotated), Backprop in Conv Layers (notes). HW1 due HW2 out 

W7: Oct 4 
Convolution as Toeplitz Matrices; Dilated/Atrous Convolution
Segmentation and Detection CNNs (and Other Pixellevel Prediction); Different Architectures. Slides, Slides (annotated). 

W8: Oct 9  Fall Student Recess  
W8: Oct 11 
Guest Lecture by Zhile Ren on 3D CNNs (PointNet, PointNet++, SPLATNet, etc). Slides 

W9: Oct 16 
De/Up/Transposed Convolution; Explainable AI. Slides, Slides (annotated). 

W9: Oct 18 
Visualizing CNNs. Slides, Slides (annotated). HW2 due HW3 out 

W10: Oct 23 
Recurrent Neural Networks. Slides, Slides (annotated), RNN notes. 

W10: Oct 25 
Recurrent Neural Networks 2 (LSTMs, RNNs + CNNs). Slides, Slides (annotated). 

W11: Oct 30 
Guest Lecture by Peter Anderson on Vision + Language (CNNs + RNNs). Slides 

W11: Nov 1 
Unsupervised Learning and Generative Modeling. Slides, Slides (annotated), Notes. 

W12: Nov 6 
Variational Autoencoders (VAEs). Slides, Slides (annotated), Notes. HW3 due 

W12: Nov 8 
(Finish) VAEs. Slides, Slides (annotated). 

W13: Nov 13  No class. CVPR Deadline.  
W13: Nov 15  No class. CVPR Deadline.  
W14: Nov 20 
Reparametrization trick and Generative Adversarial Networks (GANs). Slides, Slides (annotated). 

W14: Nov 22  No class. Thanksgiving break.  
W15: Nov 27 
Reinforcement Learning (RL) Background. Slides, Slides (annotated). 

W15: Nov 29 
Deep RL. Advanced Topics.
Slides,
Slides (annotated). 

W16: Dec 4  No class. NeurIPS. 
Grading
 80% Homework (4 homeworks)
 20% Final Project
 5% (potential bonus) Class Participation
Late policy for deliverables
 No penalties for medical reasons or emergencies. Please see GT Catalog for rules about contacting the office of the Dean of Students.
 Every student has 7 free late days (7 x 24hour chunks) for this course.
 After all free late days are used up, penalty is 25% for each additional late day.
Prerequisites
CS 4803/7643 should not be your first exposure to machine learning. Ideally, you need:
 Introlevel Machine Learning
 CS 3600 for the undergraduate section and CS 7641/ISYE 6740/CSE 6740 or equivalent for the graduate section.
 Algorithms
 Dynamic programming, basic data structures, complexity (NPhardness)
 Calculus and Linear Algebra
 positive semidefiniteness, multivariate derivates (be prepared for lots and lots of gradients!)
 Programming
 This is a demanding class in terms of programming skills.
 HWs will involve a mix of languages (Python, C++) and libraries (PyTorch).
 Your language of choice for project.
 Ability to deal with abstract mathematical concepts
FAQs

The class is full. Can I still get in?
Sorry. The course admins in IC control this process. Please talk to them.

Can I audit this class or take it pass/fail?
No. Due to the large demand for this class, we will not be allowing audits or pass/fail. Letter grades only. This is to make sure students who want to take the class for credit can.

Can I simply sit in the class (no credits)?
In general, we welcome members of the Georgia Tech community (students, staff, and/or faculty) to sitin. Out of courtesy, we would appreciate if you let us know beforehand (via email or in person). If the classroom is full, we would ask that you please allow registered students to attend.

I have a question. What is the best way to reach the course staff?
Registered students – your first point of contact is Piazza (so that other students may benefit from your questions and our answers). If you have a personal matter, email us at the class mailing list cs48037643f18staff@googlegroups.com
Project Details (20% of course grade)
The class project is meant for students to (1) gain experience implementing deep models and (2) try Deep Learning on problems that interest them. The amount of effort should be at the level of one homework assignment per group member (15 people per group).
A webpage describing the project in a selfcontained manner will be the sole deliverable. While it may link to external documents and code describing and supplementing the project, such resources may or may not be used to evaluate the project. The webpage should completely address all of the points in the rubrik described below.
Feel free to use this webpage template (zip file, hosted example) as a starting point. You do not need to follow the template, but be sure you clearly indicate how each of the sections in the rubrik below are addressed.
Submit the webpage to gradescope by uploading a zip file containing an index.html inside a project_webpage/ subdirectory (e.g. see the template). Every group member should submit this zip file and all group member names should be listed as authors on the webpage.
Rubrik (60 points)
We are not looking to see if you succeeded or failed at accomplishing what you set out to do. It’s ok if your results are not “good”. What matters is that you put in a reasonable effort, understand the project and how it related to Deep Learning in detail, and are able to clearly communicate that understanding.
A former DARPA director named George H. Heilmeier came up with a list of questions for evaluating research projects (https://www.darpa.mil/workwithus/heilmeiercatechism). We’ve adapted that list for our rubrik.
Introduction / Background / Motivation:
 (5 points) What did you try to do? What problem did you try to solve? Articulate your objectives using absolutely no jargon.
 (5 points) How is it done today, and what are the limits of current practice?
 (5 points) Who cares? If you are successful, what difference will it make?
Approach:
 (10 points) What did you do exactly? How did you solve the problem? Why did you think it would be successful? Is anything new in your approach?
 (5 points) What problems did you anticipate? What problems did you encounter? Did the very first thing you tried work?
Experiments and Results:
 (10 points) How did you measure success? What experiments were used? What were the results, both quantitative and qualitative? Did you succeed? Did you fail? Why?
In addition, 20 more points will be distributed based on presentation quality and Deep Learning knowledge:

(5 points) Appropriate use of visual aids. Are the ideas presented with appropriate illustration? Is the problem effectively visualized? Is the approach visualized appropriately? Are the results presented clearly; are the important differences illustrated? Every section and idea does not need a visual aid, but the most interesting and complex parts of the project should be illustrated.

(5 points) Overall clarity. Is the presentation clear? Can a peer who has also taken Deep Learning understand all of the points addressed above? Is sufficient detail provided?

(10 points) Finally, points will be distributed based on your understanding of how your project relates to Deep Learning. Here are some questions to think about:
 What was the structure of your problem? How did the structure of your model reflect the structure of your problem?
 What parts of your model had learned parameters (e.g., convolution layers) and what parts did not (e.g., postprocessing classifier probabilities into decisions)?
 What representations of input and output did the neural network expect? How was the data pre/postprocessed?
 What was the loss function?
 Did the model overfit? How well did the approach generalize?
 What hyperparameters did the model have? How were they chosen? How did they affect performance? What optimizer was used?
 What Deep Learning framework did you use?
 What existing code or models did you start with and what did those starting points provide?
At least some of these questions and others should be relevant to your project and should be addressed in the webpage. You do not need to address all of them in full detail. Some may be irrelevant to your project and others may be standard and thus require only a brief mention. For example, it is sufficient to simply mention the crossentropy loss was used and not provide a full description of what that is. Generally, provide enough detail that someone with an appropriate background (in both Deep Learning and your domain of choice) could replicate the main parts of your project somewhat accurately, probably missing a few less important details.
Related Classes / Online Resources
 CS231n Convolutional Neural Networks for Visual Recognition, Stanford
 Machine Learning, Oxford
 Deep Learning, New York University
 Deep Learning, CMU
 Deep Learning, University of Maryland
 Hugo Larochelle’s Neural Networks class
Book
Overviews
Note to people outside Georgia Tech
Feel free to use the slides and materials available online here. If you use our slides, an appropriate attribution is requested. Please email the instructor with any corrections or improvements.