Syllabus for CS6787

Advanced Machine Learning Systems — Fall 2018

TermFall 2018InstructorChristopher De Sa
Course[email hidden]
ScheduleMW 7:30pm – 8:45pmOffice hoursW 2:00pm – 3:00pm or by appointment
RoomUpson Hall 142OfficeBill and Melinda Gates Hall 450

Description: So you've taken a machine learning class. You know the models people use to solve their problems. You know the algorithms they use for learning. You know how to evaluate the quality of their solutions.

But when we look at a large-scale machine learning application that is deployed in practice, it's not always exactly what you learned in class. Sure, the basic models, the basic algorithms are all there. But they're modified a bit, in a bunch of different ways, to run faster and more efficiently. And these modifications are really important—they often are what make the system tractable to run on the data it needs to process.

CS6787 is a graduate-level introduction to these system-focused aspects of machine learning, covering guiding principles and commonly used techniques for scaling up to large data sets. Informally, we will cover the techniques that lie between a standard machine learning course and an efficient systems implementation. Topics will include stochastic gradient descent, acceleration, variance reduction, methods for choosing hyperparameters, parallelization within a chip and across a cluster, popular ML frameworks, and innovations in hardware architectures. An open-ended project in which students apply these techniques is a major part of the course.

Prerequisites: Knowledge of machine learning at the level of CS4780. Optionally, knowledge of computer systems and hardware on the level of CS 3410 would be useful, but this is not a prerequisite.

Format: For half of the classes, typically on Mondays, there will be a traditionally formatted lecture. For the other half of the classes, typically on Wednesdays, we will read and discuss two seminal papers relevant to the course topic. These classes will involve presentations by groups of students of the paper contents (each student will sign up in a group to present one paper for 15-20 minutes) followed by breakout discussions about the material.

Final project parameters and course calendar may be subject to change.

Grading: Students will be evaluated on the following basis.

20%Paper presentation
10%Discussion participation
30%Paper reviews — students must submit a review for every pair of papers we discuss
40%Final project

Paper review parameters: Paper reviews should be about one page (single-spaced) in length. The review guidelines should mirror what an actual conference review would look like (although you needn't assign scores or anything like that). In particular you should at least: (1) summarize the paper, (2) discuss the paper's strengths and weaknesses, and (3) discuss the paper's impact. For reference, you can read the NIPS reviewer guidelines, starting with the "Review content" section. Of course, your review will not be precisely like a real review, in large part because we already know the impact of these papers. You can submit any review up to two days late with no penalty. Students who presented a paper do not have to submit a review of that paper (although you can if you want).

Final project parameters (subject to change): The final project can be done in groups of up to three (although more work will be expected from groups with more people). The subject of the project is open-ended, but it must include:

There will be an in-class feedback activity on Wednesday, October 10, and you should prepare a two-minute pitch of your ideas by then. Project proposals are due on Monday, October 15. The project proposal should satisfy the following constraints: The project will culminate in a project report of at least four pages, not including references. The project report should be formatted similarly to a workshop paper, and should use the ICML 2018 style or a similar style. An abstract for the report is due on Monday, November 26, and we will discuss the abstracts in class on that day. The final project report is due on Wednesday, December 5.

Course Calendar

Monday, August 27 Lecture 1. [Slides] [Demo Notebook] Topics:
  • Overview
  • Course outline and syllabus
  • Gradient descent
  • Stochastic gradient descent: the workhorse of machine learning
  • Theory of SGD for convex objectives
Wednesday, August 29 In class: Sign-up for paper presentations.

Lecture 2. [Slides] [Demo Notebook] Topics:
  • The effect of choosing the step size/learning rate
  • Mini-batching and batch size
  • Overfitting
  • Generalization error
  • Early stopping
Monday, September 3 Labor day. No lecture.
Wednesday, September 5 Paper Discussion 1a. Tong Zheng. Solving large scale linear prediction problems using stochastic gradient descent algorithms. Proceedings of the twenty-first international conference on Machine learning (ICML)., 2004

Paper Discussion 1b. Rich Caruana, Steve Lawrence, and C Lee Giles. Overfitting in neural nets: Backpropagation, conjugate gradient, and early stopping. In Advances in neural information processing systems, pages 402–408, 2001
Monday, September 10 Due: Review of Paper 1a or 1b.

Lecture 3. [Slides] [Demo Notebook] Topics:
  • Regularization
  • Application-specific forms of regularization
  • The condition number
  • Momentum and acceleration
  • Momentum for quadratic optimization
  • Momentum for convex optimization
Wednesday, September 12 Due: Review of Paper 1a or 1b.

Paper Discussion 2a. Sergey Ioffe, Christian Szegedy. Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift. Proceedings of the 32nd International Conference on Machine Learning (JMLR), 2015

Paper Discussion 2b. Ilya Sutskever, James Martens, George Dahl, and Geoffrey Hinton. On the importance of initialization and momentum in deep learning. In International conference on machine learning, pages 1139–1147, 2013
Monday, September 17 Due: Review of Paper 2a or 2b.

Lecture 4. [Slides] [Demo Notebook] Topics:
  • The kernel trick
  • Gram matrix versus feature extraction: systems tradeoffs
  • Adaptive/data-dependent feature mappings
  • Dimensionality reduction
Wednesday, September 19 Paper Discussion 3a. Ali Rahimi and Benjamin Recht. Random features for large-scale kernel machines. In Advances in neural information processing systems, pages 1177–1184, 2007

Paper Discussion 3b. Kilian Weinberger, Anirban Dasgupta, Josh Attenberg, John Langford and Alex Smola. Feature Hashing for Large Scale Multitask Learning. In International conference on machine learning (ICML), 2009
Monday, September 24 Due: Review of Paper 3a or 3b.

Lecture 5. [Slides] [Demo Notebook] Topics:
  • Online versus offline learning
  • Variance reduction
  • SVRG
  • Fast linear rates for convex objectives
Wednesday, September 26 Paper Discussion 4a. Justin Ma, Lawrence K. Saul, Stefan Savage and Geoffrey M. Voelker. Identifying Suspicious URLs: An Application of Large-Scale Online Learning. In International conference on machine learning (ICML), 2009

Paper Discussion 4b. Rie Johnson and Tong Zhang. Accelerating stochastic gradient descent using predictive variance reduction. In Advances in neural information processing systems, pages 315–323, 2013
Monday, October 1 Due: Review of Paper 4a or 4b.

Lecture 6. [Slides] [Demo Notebook] Topics:
  • Hyperparameter optimization
  • Assigning parameters from folklore
  • Random search over parameters
Wednesday, October 3 Paper Discussion 5a. James Bergstra and Yoshua Bengio. Random search for hyper-parameter optimization. Journal of Machine Learning Research, 13(Feb): 281–305, 2012

Paper Discussion 5b. Jasper Snoek, Hugo Larochelle, and Ryan P Adams. Practical bayesian optimization of machine learning algorithms. In Advances in neural information processing systems, pages 2951–2959, 2012
Monday, October 8 Fall break. No lecture.
Wednesday, October 10 Due: Review of Paper 5a or 5b.

Lecture 7. [Slides] Topics:
  • Non-convex stochastic gradient descent
  • Weakness of theoretical guarantees
  • Deep learning as non-convex optimization
  • One case where we can say something: convergence to a stationary point

Activity: In-class discussion of course project ideas.
Monday, October 15 Due: Final Project Proposal.

Lecture 8. [Slides] Topics:
  • Another non-convex problem: power iteration for PCA
  • Adaptive learning rates
  • Algorithms other than SGD
Wednesday, October 17 Paper Discussion 6a. John Duchi, Elad Hazan, and Yoram Singer. Adaptive Subgradient Methods for Online Learning and Stochastic Optimization. Journal of Machine Learning Research 12.Jul (2011): 2121-2159., 2011

Paper Discussion 6b. Diederik Kingma and Jimmy Ba. Adam: A method for stochastic optimization. International Conference on Learning Representations (ICLR), 2015
Monday, October 22 Due: Review of Paper 6a or 6b.

Lecture 9. [Slides] Topics:
  • Major bottleneck for ML systems: parallelism
  • Asynchronous execution
  • Hogwild!
Wednesday, October 24 Paper Discussion 7a. Feng Niu, Benjamin Recht, Christopher Re, and Stephen Wright. Hogwild: A lock-free approach to parallelizing stochastic gradient descent. In Advances in neural information processing systems, pages 693–701, 2011

Paper Discussion 7b. Jeff Dean et al. Large scale distributed deep networks. In Advances in neural information processing systems, 2012
Monday, October 29 Due: Review of Paper 7a or 7b.

Lecture 10. [Slides] [Demo Notebook] Topics:
  • Major bottleneck for ML systems: memory bandwidth and locality
  • Low precision computation
  • Vector computation
  • Scan orders
Wednesday, October 31 Paper Discussion 8a. Suyog Gupta, Ankur Agrawal, Kailash Gopalakrishnan, and Pritish Narayanan. Deep learning with limited numerical precision. In International conference on machine learning, 2015

Paper Discussion 8b. Matthieu Courbariaux, Yoshua Bengio, and Jean-Pierre David. BinaryConnect: Training Deep Neural Networks with binary weights during propagations. In Advances in neural information processing systems, 2015
Monday, November 5 Due: Review of Paper 8a or 8b.

Lecture 11. [Slides] [Demo Notebook] Topics:
  • What happens on the inference side?
  • Specialized low-cost models
  • Compression
  • Hardware for machine learning
  • The dominance of GPUs
  • Accelerators for machine learning
  • Will all computation become matrix multiply?
Wednesday, November 7 Paper Discussion 9a. Wenlin Chen, James Wilson, Stephen Tyree, Kilian Weinberger, and Yixin Chen. Compressing Neural Networks with the Hashing Trick. Proceedings of the 32nd International Conference on Machine Learning (ICML), 2015

Paper Discussion 9b. Song Han, Huizi Mao, and William J Dally. Deep compression: Compressing deep neural networks with pruning, trained quantization and Huffman coding. ICLR, 2016
Monday, November 12 Due: Review of Paper 9a or 9b.

Lecture 12. [Slides] Topics:
  • Machine learning frameworks
  • TensorFlow
  • SciKit-Learn
  • PyTorch
  • Is Python the ML language of the future?
Wednesday, November 14 Paper Discussion 10a. Norman P Jouppi, Cliff Young, Nishant Patil, David Patterson, Gaurav Agrawal, Raminder Bajwa, Sarah Bates, Suresh Bhatia, Nan Boden, Al Borchers, et al. In-datacenter performance analysis of a tensor processing unit. In Proceedings of the 44th Annual International Symposium on Computer Architecture, pages 1–12. ACM, 2017

Paper Discussion 10b. Martin Abadi et al. TensorFlow: A System for Large-Scale Machine Learning. USENIX Symposium on Operating Systems Design and Implementation (OSDI), 2016
Monday, November 19 Due: Review of Paper 10a or 10b.

Lecture 13. [Slides] [Demo Notebook] Topics:
  • Final project abstract and report requirements
  • Sparse matrix computations
  • Structured matrices
Wednesday, November 21 Thanksgiving break. No lecture.
Monday, November 26 Due: Abstract for Final Project.

Abstract swap and discussion.
Wednesday, November 28 Abstract swap continued.

Lecture 14. Topics: discussion of advances in ML accelerator hardware.
Monday, December 3 Lecture 15. Epilogue.
  • Recap of course content
Wednesday, December 5 No lecture. Due: Final Project Report.