CS 342 - Neural networks - Fall 2018

TTh 10am-11am UTC 3.124
Fr 10am-11am / 11pm-noon ECJ 1.204

instructor Philipp Krähenbühl ( philkr (at) utexas.edu )
office hours Tue 11am-12pm GDC 4.824
Please only visit during office hours or with appointment (or if I didn’t reply your email after the third try)

Dian Chen ( dchen (at) cs.utexas.edu )
office hours Mon 11am-noon GDC 1.302 Desk 1
Ankur Garg ( ankgarg (at) cs.utexas.edu )
office hours Mon 4pm-5pm GDC 1.302 Desk 1

We will use piazza for questions and canvas homework.


class section

Date Topic Concepts Slides Corresponding book chapters (optional; see below) Notes and due dates
Aug 30 Introduction tensors layers intro Background: Ch 1-5  
Aug 31 Installing pytorch and building the first compute graph   slides
  HW1 out
Sep 4 Simple networks fully connected ReLU slides exercise 1 solution    
Sep 6 Output transformations and loss functions sigmoid softmax log likelihood L1/L2 loss slides 5.9 HW1 due 11:59pm solution
Sep 7 Logistic regression and MLP   slides
  HW2 out
Sep 11 Back propagation back prop slides exercise 2 solution Ch 6  
Sep 13 Stochastic gradient descent SGD momentum slides   HW2 due 11:59pm solution
Sep 14 Regression vs classification   slides   HW3 out
Sep 18 Activation functions + Initialization LeakyReLU PReLU Gaussian Init Xavier Init slides Ch 6  
Sep 20 Convolutions + Pooling Conv2D Avg Pooling Max Pooling striding / padding slides 9.1-9.3 HW3 due 11:59pm solution
Sep 21 ConvNets   slides
  HW4 out
Sep 25 Overfitting early stopping data augmentation dropout parameter sharing ensembles slides exercise 3 solution 3 7.4, 7.8, 7.9, 7.11, 7.12  
Sep 27 Optimization tips and tricks weight regularization batch normalization slides exercise 4 solution 4 7.1-7.3 HW4 due 11:59pm solution
Sep 28 Regularization and normalization   slides
  HW5 out
Oct 2 Popular Architectures I LeNet AlexNet VGG auxillary losses slides exercise 5 solution    
Oct 4 Popular Architectures II GoogLeNet ResNet DenseNet residual connections slides   HW5 due 11:59pm solution
Oct 5 ResNets   slides   HW6 out
Oct 9 Object detection ROI Pooling R-CNN Fast R-CNN YOLO slides exercise 6 solution    
Oct 11 Segmentation up-convolution FCN slides   HW6 due 11:59pm solution
Oct 12 Fully convolutional networks   slides
  HW7 out
Oct 16 Visualization and understanding gradients and saliency slides exercise 7 solution 7    
Oct 18 Generative models reparametrization trick adversarial loss slides 20.10, 20.12, 20.13 HW7 due 11:59pm solution
Oct 19 Upconvolution and deep rendering   slides
  HW8 out
Oct 23 Recurrent Network I RNN LSTM GRU slides exercise 8 solution 8 10.1-10.11  
Oct 25 Recurrent Network II Temporal CNNs slides   HW8 due 11:59pm solution
Oct 26 Future prediction   slides
  HW9 out
Oct 30 Deep Reinforcement Learning I policy gradient slides exercise 9 solution 9    
Nov 1 Deep Reinforcement Learning II Q-learning actor-critic slides   HW9 due 11:59pm solution1 solution2
Nov 2 Immitation learning   slides   HW10 out
Nov 6 Beyond RL I gradient free optimization slides exercise 10 solution 10    
Nov 8 Beyond RL II imitation learning dagger direct future prediction slides   HW10 due 11:59pm solution
Nov 9 Gradient free optimization   slides   HW11 out
Nov 13 Embedding learning triplet loss contrastive loss slides exercise 11 solution 11    
Nov 15 Style transfer gram statistic slides    
Nov 16 Driving in SuperTuxKart       HW11 due 11:59pm
Final project out
Nov 20 Language   slides    
Nov 22 no class        
Nov 23 no class        
Nov 27 Adversarial examples adversarial attack slides 7.13  
Nov 29 Negative results        
Nov 30 Final project Q/A        
Dec 4 Final project flash presentations       Final project due 11:59pm
Dec 6 Final project competition        
Dec 7 Nagging TA for better grade       no class

concepts: layers/architecture algorithm supervision applications


Class overview

Goals of the class

After this class you should be able to


There will be in class quizzes throughout the semester (about one a week). You can work with your peers on those quizzes during class. Each quiz will be graded pass/fail.

There will be a homework assignment every week, but we will keep them small and manageable. Homework needs to be solved individually, if we detect duplicate solutions you will lose the full credit for the assignment. Most of the homework will be graded automatically, you’ll have access to a partial grader while you’re working on the assignment. We will use the same grader, but different test data for your final grade. You’ll submit your homework through canvas.

Late policy:

Expected workload

Estimates of required effort to pass the class are:

Course material

The course should be self contained, but if you need additional reading material consult Deep learning, Goodfellow, Bengio and Courville 2016 www.deeplearningbook.org.


Syllabus subject to change.