Syllabus

Please find all the Video Lectures within this YouTube Playlist.

Lecture Slides are password protected. 

• Week 01: Introduction, course description, administrative details, features spaces, k-nearest neighbors, classification
  • Lecture Slides
• Week 02: Machine Learning Pipeline, Linear classifiers, Classifier Evaluation,
  • Lecture Slides
• Week 03: Loss Functions, Optimization, Gradient Descent
  • Lecture Slides (Stanford CS231n Lecture 3 Slides)
  • Optimization Demo (Stanford CS231n)
• Week 04: Neural Networks, Perceptron Layers, Backpropagation
  • Lecture Slides (Stanford CS231n Lecture 4 Slides)
• Week 05: Convolutional, Pooling and Soft-Max Layers, Convolutional Neural Networks (CNN)
  • Lecture Slides (password protected)
  • and Stanford CS231n Lecture 5 Slides are also very good.
  • Here are the layers of AlexNet, we will go over it in class.
  • and finally Homework 1(due November 8th) 
• Week 06: Analysis of a complete Network: AlexNet and a practical session on deep learning coding
  • Practicum: Using the (password proceted) code, we will train our first classification network. This piece of code will be your base for your projects. 
• Week 07: Training CNNs, activation functions, initialization, batch normalization...
  • Lecture Slides (Stanford CS231n Lecture 6 Slides)
  • Here is a good article on activations functions.
  • And here is another nice article on data preprocessing (whitenning).
  • also, this is another nice article on weight initialization.
• Week 08: More on Training, Solvers, Regularization Techniques, Data Augmentation, Transfer Learning.
  • Lecture Slides (Stanford CS231n Lecture 7 Slides)
  • Here is a useful article on optimization algorithms (solvers) in CNNs.
• Week 09: Classification Architectures: AlexNet, VGG, GoogLeNet, ResNet...
  • Lecture Slides (Stanford CS231n Lecture 9 Slides)
• Week 10: Analysis on a complete Deep Learning Project
  • (dataset construction, training, results analysis etc.)
  • here is some codes.
• Week 11: Introduction to Recurrent Neural Networks
  • Lecture Slides (Stanford CS231n Lecture 10 Slides)
  • Those who are interested can find a very nice (both mathematical and philosophical) introduction to the subject here.
• Week 12:This is the midterm week.
• Week 13: Memory, GRU and LSTM, Advanced CNN architectures.
  • Here are the slides (password protected.)
  • Nice article on GRU and LSTMs. Watch its YouTube video as well.
• Week 14: Project Presentations