| 1/29 |
Thu |
No Class. Class begins on Tuesday 2/3. |
|
| 1. 2/3 |
Tue |
Course overview. Probability review. Linearity of expectation and variance. |
Slides. Compressed slides. Reading: MIT short videos and exercises on probability (go to Unit 4). Khan academy probability lessons (a bit more basic). Chapters 1-3 of Probability and Computing with content and excersises on basic probability, expectation, variance, and concentration bounds. |
| Randomized Methods, Sketching & Streaming |
| 2. 2/5 |
Thu |
Estimating set size by counting duplicates. Markov's inequality. Random hashing for efficient lookup. Collision-free hashing. |
Slides. Compressed slides. Reading: Chapters 1-3 of Probability and Computing with content and excersises on basic probability, expectation, variance, and concentration bounds. |
| 3. 2/10 |
Tue |
2-level hashing. 2-universal and pairwise independent hashing. |
Slides. Compressed slides. Reading: Chapter 2.2 of Foundations of Data Science with content on Markov's inequality and Chebyshev's inequality. Exercises 2.1-2.6. Chapters 1-3 of Probability and Computing with content and excersises on basic probability, expectation, variance, and concentration bounds. Some notes (Arora and Kothari at Princeton) proving that the ax+b mod p hash function described in class in 2-universal.
|
| 4. 2/12 |
Thu |
Hashing for load balancing and Chebyshev's inequality. The union bound. Motivate exponential concentration bounds. |
Slides. Compressed slides. Reading: Chapter 2.2 of Foundations of Data Science with content on Markov's inequality and Chebyshev's inequality. Exercises 2.1-2.6. Chapters 1-3 of Probability and Computing with content and excersises on basic probability, expectation, variance, and concentration bounds.
|
| 5. 2/17 |
Tue |
Class held over Zoom. Exponential concentration bounds and the central limit theorem. |
Slides. Reading: Chapter 4 of Probability and Computing on exponential concentration bounds. Some notes (Goemans at MIT) showing how to prove exponential tail bounds using the moment generating function + Markov's inequality approach. |
| 2/19 |
Thu |
No Class. Monday class schedule followed. |
|
| 6. 2/24 |
Tue |
Finish up applications of exponential concentration bounds. Bloom Filters. |
Slides. Reading: Chapter 4 of Mining of Massive Datasets, with content on Bloom filters. See here for full Bloom filter analysis. See Wikipedia for a discussion of the many bloom filter variants, including counting Bloom filters, and Bloom filters with deletions. |
| 7. 2/26 |
Thu |
Finish up Bloom filters. Start on streaming algorithms and frequent elements estimation. |
Slides. Reading: Chapter 4 of Mining of Massive Datasets, with content on Bloom filters. Notes (Amit Chakrabarti at Dartmouth) on streaming algorithms. See Chapters 1 and 5 for frequent elements. Some more notes on the frequent elements problem. |
| 8. 3/3 |
Tue |
Frequent elements estimation via Count-min sketch. Min-Hashing for Distinct elements. |
Slides. Reading: Notes (Amit Chakrabarti at Dartmouth) on streaming algorithms. See Chapters 1 and 5 for frequent elements. Some more notes on the frequent elements problem. A website with lots of resources, implementations, and example applications of count-min sketch. Chapter 4 of Mining of Massive Datasets, with content on distinct elements counting.
|
| 9. 3/5 |
Thu |
Finish up distinct elements counting. The median trick. Distinct elements in pratice: Flajolet-Martin and HyperLogLog. |
Slides. Reading: Chapter 4 of Mining of Massive Datasets, with content on distinct elements counting. The 2007 paper introducing the popular HyperLogLog distinct elements algorithm.
|
| 10. 3/10 |
Tue |
Jaccard similarity, fast similarity search, and locality sensitive hashing |
Slides. Reading: Chapter 3 of Mining of Massive Datasets, with content on Jaccard similarity, MinHash, and locality sensitive hashing.
|
| 3/12 |
Thu |
Midterm 1. 1-2:15pm. In class. |
Study guide and review questions. |
| 3/17 |
Tue |
No Class. Spring Break. |
|
| 3/19 |
Thu |
No Class. Spring Break. |
|
| Spectral Methods |
| 12. 3/24 |
Tue |
Compressing high dimensional data: low-distortion embeddings and the Johnson-Lindenstrauss Lemma. Example application to clustering.
|
Slides. Reading: Chapter 2.7 of Foundations of Data Science on the Johnson-Lindenstrauss lemma. Notes on the JL-Lemma (Anupam Gupta (CMU). Sparse random projections which can be multiplied by more quickly. Some good videos for linear algebra review.. See also: Khan academy. |
| 13. 3/26 |
Thu |
Intro to principal component analysis, low-rank approximation, data-dependent dimensionality reduction. Orthogonal bases and projection matrices. Dual column/row view of low-rank approximation. |
Slides.
Reading: Chapter 3 of Foundations of Data Science and Chapter 11 of Mining of Massive Datasets on low-rank approximation and the SVD. Some good videos overviewing the SVD and related topics (like orthogonal projection and low-rank approximation). |
14. 3/31 |
Tue |
Best fit subspaces and optimal low-rank approximation via eigendecomposition. |
Slides. Reading: Proof that optimal low-rank approximation can be found greedily (see Section 1.1). Chapter 3 of Foundations of Data Science and Chapter 11 of Mining of Massive Datasets on low-rank approximation. |
| 15. 4/2 |
Thu |
Finish up optimal low-rank approximation via eigendecomposition. Eigenvalues as a measure of low-rank approximation error. General linear algebra review.
|
Slides.
Reading: Chapter 3 of Foundations of Data Science and Chapter 11 of Mining of Massive Datasets on low-rank approximation. |
| 16. 4/7 |
Tue |
The singular value decomposition and connections to low-rank approximation. Applications of low-rank approximation beyond compression. Matrix completion and entity embeddings.
|
Slides.
Reading: Notes on SVD and its connection to eigendecomposition/PCA (Roughgarden and Valiant at Stanford). Levy Goldberg paper on word embeddings as implicit low-rank approximation. |
| 17. 4/9 |
Thu |
Spectral graph theory and spectral clustering. |
Slides. Reading: Chapter 10.4 of Mining of Massive Datasets on spectral graph partitioning. For a lot more interesting material on spectral graph methods see Dan Spielman's lecture notes. Great notes on spectral graph methods (Roughgarden and Valiant at Stanford). |
| 18. 4/14 |
Tue |
The stochastic block model. |
Slides. Reading: Dan Spielman's lecture notes on stochastic block model, including matrix concentration + David-Kahan perturbation analysis.. Further stochastic block model notes (Alessandro Rinaldo at CMU). A survey of the vast literature on the stochastic block model, beyond the spectral methods discussed in class (Emmanuel Abbe at Princeton). |
| 19. 4/16 |
Thu |
Computing the SVD: power method. |
Slides. Reading: Chapter 3.7 of Foundations of Data Science on the power method for SVD. Some notes on the power method. (Roughgarden and Valiant at Stanford). |
| 20. 4/21 |
Tue |
Finish up power method analysis. Krylov methods. Connection to random walks and Markov chains. |
Slides. Reading: Chapter 3.7 of Foundations of Data Science on the power method for SVD. Some notes on the power method. (Roughgarden and Valiant at Stanford). Multivariable calc review, e.g., through: Khan academy |
| 4/23 |
Thu |
Midterm 2. 1-2:15pm. In class. |
Study guide and review questions. |
| Optimization |
| 21. 4/28 |
Tue |
Intro to gradient descent and its analysis for convex Lipschitz functions. |
Slides. Reading: Chapters I and III of these notes (Hardt at Berkeley). |
| 22. 4/30 |
Thu |
Finish gradient descent analysis. Constrained optimization and projected gradient descent. Start on motivation for stochastic gradient descent. |
Slides. Reading: Short notes, proving regret bound for online gradient descent. A good book (by Elad Hazan) on online optimization, including online gradient descent and connection to stochastic gradient descent |
| 23. 5/4 |
Tue |
Online gradient descent and application to the analysis of stochastic gradient descent. |
Slides. Reading: Short notes, proving regret bound for online gradient descent. A good book (by Elad Hazan) on online optimization, including online gradient descent and connection to stochastic gradient descent. |
| 24. 5/7 |
Thu |
Course wrap-up and final exam review. |
Slides. |
| 5/12 |
Tue |
Final Exam. 1-3pm. In class. |
Study guide and review questions. |