COMPSCI 682 Neural Networks: A Modern Introduction


  • This is a tentative class outline and is subject to change throughout the semester.
  • Slides will be finalized after each lecture.
Event TypeDateDescriptionCourse Materials
Lecture Tuesday, Sep 5 Intro to deep learning, Historical context [lecture recording]
[python/numpy tutorial]
[software setup for assignments]
Lecture Thursday, Sep 7 Image classification and the data-driven approach
k-nearest neighbor
Linear classification
[lecture recording]
[image classification notes]
[linear classification notes]
Lecture Tuesday, Sep 12 Loss functions
[lecture recording]
Lecture Thursday, Sep 14 Optimization
Stochastic gradient descent
[lecture recording]
[backprop notes]
[Efficient BackProp] (optional)
related: [1], [2], [3] (optional)
Optional discussion Friday, Sept. 15 Python setup, Google collab, Basics of Python and NumPY
Lecture Tuesday, Sep 19 Backpropagation
Vector, Matrix, and Tensor Derivatives
[lecture recording]
handout 1: Vector, Matrix, and Tensor Derivatives
handout 2: Derivatives, Backpropagation, and Vectorization
Deep Learning [Nature] (optional)
Lecture Thursday, Sep 21 Neural networks
Training neural networks
Activation functions
[lecture recording]
[Neural Nets notes 1]
tips/tricks: [1], [2] (optional)
Optional discussion Friday, Sept. 22 Reviewing the chain rule, Applying the chain rule to vectors [slides]
Lecture Tuesday, Sep 26 Training neural networks II
Weight initialization
Batch normalization
[lecture recording]
[Neural Nets notes 2]
[Batch Norm]
Lecture Thursday, Sep 28 Training neural networks III
Babysitting the learning process
Hyperparameter optimization
[lecture recording]
[Bengio 2012] (optional)
Optional discussion Friday, Sept. 29 Vector, Matrix, and Tensor Derivatives [slides]
Lecture Tuesday, Oct 3 Project discussion led by TAs
Expectations and timeline
Overview of projects outlined by the TAs
Thursday, Oct 5 Instructors at ICCV, no class
Tuesday, Oct 10 Monday schedule, no class
Lecture Thursday, Oct 12 Training neural networks IV
Model ensembles, dropout
Convolutional neural networks
[lecture recording]
Optional discussion Friday, Oct. 13 Batch normalization [slides]
Lecture Tuesday, Oct 17 Convolutional neural networks
[lecture recording]
ResNet (optional)
FCN (optional)
Lecture Thursday, Oct 19 ConvNets for spatial localization I
Object detection
[lecture recording]
Lecture Tuesday, Oct 24 ConvNets for spatial localization II
Image Segmentation
[lecture recording]
Lecture Thursday, Oct 26 Visualizing and understanding ConvNets [lecture recording]
[visualization notes]
Lecture Tuesday, Oct 31 Neural Texture Synthesis and Style Transfer [lecture recording]
Lecture Thursday, Nov 2 Guest lecture: James Tompkin, MLFL (12-1pm, CS 150)
More Cameras and Better Cameras for Scene Reconstruction
[talk info]
Lecture Tuesday, Nov 7 Guest lecture: Sajjad Amini
Generative modeling: Autoregressive, Variational, GANs
[lecture recording]
[GAN Notes]
Lecture Thursday, Nov 9 Recurrent neural networks [lecture recording]
Lecture Tuesday, Nov 14 Midterm review led by TAs [lecture recording]
Midterm Thursday, Nov 16 Midterm to be held during regular lecture time
Lecture Tuesday, Nov 21 Transformers [lecture recording]
Lecture Thursday, Nov 23 Thanksgiving, no class
Lecture Tuesday, Nov 28 Multi-modal AI [lecture recording]
Lecture Thursday, Nov 30 Guest lecture: Sara Beery, MLFL (12-1pm, CS150)
Monitoring the Urban Forest with Auto Arborist
Lecture Tuesday, Dec 5 Neuro-Symbolic AI [lecture recording]
Lecture Thursday, Dec 7 Self-supervised Learning