csc311
CSC 311 Spring 2020: Introduction to Machine Learning
Machine learning (ML) is a set of techniques that allow computers to learn from data and experience, rather than requiring humans to specify the desired behaviour manually. ML has become increasingly central both in AI as an academic field, and in industry. This course introduces the main concepts and ideas in machine learning, and provides an overview of many commonly used machine learning algorithms. It also serves as a foundation for more advanced ML courses.
By the end of this course, the students will learn about (roughly categorized)
- Machine Learning Problems: Supervised (regression and classification), Unsupervised (clustering, dimension reduction), Reinforcement Learning
- Models: Linear and Nonlinear (Basis Expansion and Neural Networks)
- Loss functions: Squared Loss, Cross Entropy, Hinge, Exponential, etc.
- Regularizers: l1 and l2
- Probabilistic viewpoint: Maximum Likelihood Estimation, Maximum A Posteriori, Bayesian inference
- Bias and Variance Tradeoff
- Ensemble methods: Bagging and Boosting
- Optimization technique in ML: Gradient Descent and Stochastic Gradient Descent
The students are expected to learn the intuition behind many machine learning algorithms and the mathematics behind them. Through homework assignments, they will also learn how to implement these methods and use them to solve simple machine learning problems. More details can be found in syllabus and piazza.
Announcements:
- Midterm exam detail:
- The main exam is on February 24th (Monday) at 6PM. The locations are:
- Section L0101: Muzzo Family Alumni Hall, AH 100
- Section L5101: Brennan Hall, BR 200
- The makeup exam is on February 24th (Monday) at 8PM. Its location is at Bahen, BA 2139. Note: If you want to take the makeup exam, you have to contact us and get the permission.
- Topics are everything up to Multi-class Classification, but not Neural Networks.
- The main exam is on February 24th (Monday) at 6PM. The locations are:
- The final exam is cancelled because of the COVID-19 pandemic, so we have a new marking scheme:
- Four (4) homework assignments (12% each): 48% (increased from 40%)
- Five (5) reading assignments: 10%
- Take-home test: 12% (the new component)
- Midterm: 30% (increased from 20%)
- The take-home test will be released on April 18th (Saturday) morning. You have until April 20th (Monday) at 5PM to upload it on MarkUs. We won’t release the test on this website; please check Quercus and/or Piazza for them.
Instructors:
| Prof | Amir-massoud Farahmand | Emad A. M. Andrews |
|---|---|---|
| csc311-2020-01@cs.toronto.edu | csc311-2020-01@cs.toronto.edu | |
| Office hours | Thursday, 13-14 at BA2283 | Thursday, 20-22 at BA2283 |
Teaching Assistants:
Chunhao Chang, Rasa Hosseinzadeh, Julyan Keller-Baruch, Navid Korhani, Shun Liao, Ehsan Mehralian, Alexey Strokach, Jade Yu
Time & Location:
| Section | Room | Lecture time | Tutorial time |
|---|---|---|---|
| L0101 | MS 2172 | Tuesday 13-15 | Thursday 14-15 |
| L5101 | BA 1130 | Thursday 18-20 | Thursday 20-21 |
Suggested Reading
No required textbooks. Suggested reading will be posted after each lecture (See lectures below).
- (ESL) Trevor Hastie, Robert Tibshirani, and Jerome Friedman, The Elements of Statistical Learning, 2009.
- (PRML) Christopher M. Bishop, Pattern Recognition and Machine Learning, 2006.
- (RL) Richard S. Sutton and Andrew G. Barto Reinforcement Learning: An Introduction, 2018.
- (DL) Ian Goodfellow, Yoshua Bengio and Aaron Courville (2016), Deep Learning
- (MLPP) Kevin P. Murphy, Machine Learning: A Probabilistic Perspective, 2013.
- (ISL) Gareth James, Daniela Witten, Trevor Hastie, and Robert Tibshirani, Introduction to Statistical Learning, 2017.
- () Shai Shalev-Shwartz and Shai Ben-David Understanding Machine Learning: From Theory to Algorithms, 2014.
- (ITIL) David MacKay, Information Theory, Inference, and Learning Algorithms, 2003.
Lectures
This is the tentative schedule, and it may change.
| Week (date) | Topics | Lectures | Tutorial materials | Suggested reading |
|---|---|---|---|---|
| 1 (Jan 6) |
Introduction to ML & Nearest Neighbours | slides | Probability review slides & preliminaries | ESL 1. ESL 2.(1-3), 2.5 |
| 2 (Jan 13) |
Decision Trees | slides & Example | Linear Algebra review slides | ESL 9.2 |
| 3 (Jan 20) |
Linear regression Optimization |
slides | Optimization and CV slides notebook worksheet | PRML 3.2 ESL 2.9, 8.7 ESL 3.2 |
| 4 (Jan |
Bias-Variance Decomposition Ensembles: Bagging Linear Methods for Classification |
slides | No tutorial | PRML 4.3 ESL 4.(1,2,4) |
| 5 (Feb |
(Continued) Bias-Variance Decomposition Ensembles: Bagging Linear Methods for Classification |
Gaussian distribution; Bias-variance Optimization for logistic regression slides |
[notes] | |
| 6 (Feb |
Multiclass Classification Neural Networks |
slides | Midterm review slides | notes |
| 7 (Feb |
Support Vector Machines | slides | Invited Speaker, Terrance D’souza, 8:00pm at BA 1130 | ESL 12.(1-3) |
| 8 (Mar |
Ensembles: Boosting | slides | Deep Learning with PyTorch (three modules covered) | ESL 10.(1-5) |
| 9 (Mar |
Probabilistic models | slides | More on SVM and Kernel Methods | ESL 2.6.3, 6.6.3, 4.3.0 |
| 10 (Mar |
(Continued) Probabilistic models | Videos: Farahmand: MLE & Naive Bayes, Bayesian (Part 1), Bayesian (Part 2), GDA Andrews: Probabilistic Models |
PRML 12.1 | |
| 11 (Mar |
Principal Component Analysis | slides | Videos: Farahmand: PCA (Part 1), PCA (Part 2) Andrews: PCA Matrix Factorizations & Recommender Systems More on k-Means & EM Algorithm & EM demo on Kaggle More on Recommender Systems & Team Dinosaur Planet |
ESL 14.5.1 |
| 12 (Mar |
k-Means & EM Algorithm Reinforcement Learning |
KMeans slides RL slides |
Videos: Farahmand: Clustering (Part 1), Clustering (Part 2), RL (Part 1), RL (Part 2), RL (Part 3) Andrews: Clustering |
PRML 9.1-2 RL 3, 4.1, 4.4, 6.1-6.5 |
Homework Assignments
This is a tentative schedule of the homework assignments. Most of them will be released on Thursday evening and they will be due in 10 days. Please see the course information handout for detailed policies (marking, lateness, etc.).
| Homework # | Out | Due | Materials | TA Office Hours |
|---|---|---|---|---|
| Homework 1 | Jan 23, 23:59 | Feb 3, 16:59 | Questions Dataset | Jan 28 (Tuesday), 4-6PM at BA3289 Jan 31 (Friday), 3-5PM at BA3289 |
| Homework 2 | Feb 6, 23:59 | Feb 17, 16:59 | Questions Dataset | Feb 11 (Tuesday), 4:30-6:30PM at BA3289 Feb 14 (Friday), 3-5PM at BA3289 |
| Homework 3 | Mar 5, 23:59 | Mar 16, 16:59 | Questions Dataset | Mar 10 (Tuesday), 4-6PM at BA3289 Mar 13 (Friday), 11:30AM-1PM at BA4290 |
| Homework 4 | Mar 23, 23:59 | Apr 3, 16:59 | Questions Dataset | Mar 27 (Friday), 2-4PM at virtual office Mar 31 (Tuesday), 4-6PM at virtual office |
Reading Assignments
There will be five (5) reading assignments. These are selected from seminal papers in machine learning or closely related fields such as statistics. These papers complement the topics we cover in the course, or show you how a research paper in ML is written. You have to read them and try to understand as much as possible. Some papers are easy and some may be difficult. It is not important that you completely understand a paper, but you should put some effort into it.
We ask you to summarize each paper in a short paragraph (100-200 words) and try to come up with two suggestions on how the method described in the paper can be used or extended. These five assignments contribute 10% to your final mark.
The reading assignments are only lightly evaluated: You should submit your summary before April 10th (Friday) at 5PM. We randomly check some of your summaries to see whether or not they are in fact relevant to the assigned paper. If they are, you get the points. If they are not, and you submitted a filler, you will get only 0% (even if it is only one of your submissions).
We will post the papers as the course progresses.
- Robert Tibshirani, “Regression Shrinkage and Selection Via Lasso,” Journal of the Royal Statistical Society, 1996. PDF
- Leon Bottou and Olivier Bousquet, “The Tradeoffs of Large Scale Learning,” Advances in Neural Information Processing Systems (NeurIPS), 2007. PDF
- Alex Krizhevsky, Ilya Sutskever, and Geoffrey E. Hinton, “Imagenet classification with deep convolutional neural networks,” Advances in Neural Information Processing Systems (NeurIPS), 2012. PDF
- Moritz Hardt, Eric Price, and Nathan Srebro, “Equality of opportunity in supervised learning,” Advances in Neural Information Processing Systems (NeurIPS), 2016. PDF; Long version - Optional
- Volodymyr Mnih, Koray Kavukcuoglu, David Silver, et al., “Human-level control through deep reinforcement learning,” Nature, 2015. PDF
Computing Resources
For the homework assignments, we will use Python, and libraries such as NumPy, SciPy, and scikit-learn. You have two options:
-
The easiest option is probably to install everything yourself on your own machine.
-
If you don’t already have python, install it. We recommend using Anaconda. You can also install python directly if you know how.
- Optionally, create a virtual environment for this class and step into it. If you have a conda distribution run the following commands:
conda create --name csc311 source activate csc311 - Use pip to install the required packages
pip install scipy numpy autograd matplotlib jupyter sklearn
-
-
All the required packages are already installed on the Teaching Labs machines.