| Week | Content | Deliverables Due |
|---|---|---|
| 1 | Nearest Neighbors Bias-Variance Trade-Off Cross-Validation |
|
| 2 | Classification and regression trees | |
| 3 | Bagging and boosting | HW 1 |
| 4 | Evaluating classifiers | |
| 5 | Support Vector Machines | HW 2 |
| 6 | Intro to Neural networks | Individual HW |
| 7 | Deep learning | Midterm |
| 8 | Recommender Systems | |
| 9 | Sequence Modelling | HW 3 |
| 10 | In-class project presentation | |
| 11 | No class; Project write-up due | HW 4 |
Week 9
Sequence Modelling — First half of the lecture
Guest speaker: Vinh Luong — Second half of the lecture
Code: script
Optional reading:
- Deep Learning by Ian Goodfellow and Yoshua Bengio and Aaron Courville
Book available at: http://www.deeplearningbook.org/
Chapter 15
Week 8
- recommender systems
Recommender Systems — Guest lecture: Rina Foygel Barber
Code: script
Optional reading:
- We will be using recommenderlab package.
- Amazon.com Recommendations
- Cold Start Problem — Finding a Needle in a Haystack of Reviews
- Matrix Factorization Techniques For Recommender Systems
- All Together Now: A Perspective on the Netflix Prize
Week 7
- convolutional neural networks
- auto-encoders
Optional reading:
- Deep Learning by Ian Goodfellow and Yoshua Bengio and Aaron Courville
Book available at: http://www.deeplearningbook.org/
Chapter 9 and Chapter 14 provide enormous amount of detail.
Week 6
- Introduction to neural networks
Neural networks
Code: scripts
Optional reading:
-
Deep Learning by Ian Goodfellow and Yoshua Bengio and Aaron Courville
Book available at: http://www.deeplearningbook.org/
Read chapter 6. Skim through chapters 7 and 8. - h2o booklet on deep learning
- h2o package
Week 5
- perceptron, linear classifiers
- support vector machines
- gradient descent
SVM,
Optimization
Code: scripts
Optional reading:
- ISLR - Section 9
Individual homework: PDF
Week 4
- evaluating classifiers; confusion matrix
- expected value framework
- profit and lift curves
Optional reading:
- ISLR - Section 4
Week 3
- ensemble methods
- bagging; random forest
- boosting
Code:
Optional reading:
Week 2
Slides
Code:
Optional reading:
- ISLR - Section 4, Section 8
- Overview of trees PDF
Week 1
- Introduction to supervised learning; regression and classification
- k nearest neighbors methods; similarity in machine learning
- bias-variance trade-off; cross-validation
Code:
- Boston Housing R
- bias-variance-illustration.R
Optional textbook reading: An Introduction to Statistical Learning: Section 2, Section 5.1
- PDF can be freely obtained here.
Instructions on how to convert the R Markdown document (Rmd file) into a PDF file.
Assignement 1 Due 11.59pm on Friday, January 19