Lecture      | Day | Topic | Materials/Reading |
---|---|---|---|
1. 9/6 | Tue | Course overview. Probability review. Linearity of expectation. | Slides. Compressed slides. MIT short videos and exercises on probability. Khan academy probability lessons (a bit more basic). |
Randomized Methods, Sketching & Streaming | |||
2. 9/8 | Thu | Estimating set size by counting duplicates. Concentration Bounds: Markov's inequality. Random hashing for efficient lookup. | Slides. Compressed slides. |
3. 9/13 | Tue | Analysis of random hashing. 2-level hashing. Markov's ineqaulity. | 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. |
4. 9/15 | Thu | 2-universal and pairwise independent hashing. Hashing for load balancing. Chebyshev's inequality. The union bound. | Slides. Compressed slides. Some notes (Arora and Kothari at Princeton) proving that the ax+b mod p hash function described in class in 2-universal. |
5. 9/20 | Tue | Exponential concentration bounds and the central limit theorem. | Slides. Compressed slides. Reading: Some notes (Goemans at MIT) showing how to prove exponential tail bounds using the moment generating function + Markov's inequality approach discussed in class. |
6. 9/22 | Thu | Finish up exponential concentration bound. Bloom Filters. | Slides. Compressed slides. Reading: Chapter 4 of Mining of Massive Datasets, with content on bloom filters. See here for full Bloom filter analysis. See here for some explaination of why a version of a Bloom filter with no false negatives cannot be achieved without using a lot of space. See Wikipedia for a discussion of the many bloom filter variants, including counting Bloom filters, and Bloom filters with deletions. See Wikipedia again and these notes for an explaination of Cuckoo Hashing, a randomized hash table scheme which, like 2-level hashing, has O(1) query time, but also has expected O(1) insertion time. |
7. 9/27 | Tue | Finish up Bloom Filters. Min-Hashing for Distinct Elements. | Slides. Compressed slides. Reading: Chapter 4 of Mining of Massive Datasets, with content on distinct elements counting. |
8. 9/29 | Thu | Distinct elements analysis. The median trick. Distinct elements in pratice: Flajolet-Martin and HyperLogLog. | Slides. Compressed 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. |
9. 10/4 | Tue | Jaccard similarity estimation with MinHash. Locality sensitive hashing for fast similarity search. | Slides. Compressed slides. Reading: Chapter 3 of Mining of Massive Datasets, with content on Jaccard similarity, MinHash, and locality sensitive hashing. |
10. 10/6 | Thu | Finish up LSH. The frequent elements problem and count-min sketch. | Slides. Compressed slides. Reading: Notes (Amit Chakrabarti at Dartmouth) on streaming algorithms. See Chapters 2 and 4 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. |
11. 10/11 | Tue | Finish up frequent elements estimation. Dimensionality reduction, low-distortion embeddings, and the Johnson Lindenstrauss Lemma. | Slides. Compressed 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. Linear Algebra Review: Khan academy. |
12. 10/13 | Thu | Johnson-Lindenstrauss lemma proof. | Slides. Compressed slides. Reading: Chapter 2.7 of Foundations of Data Science on the Johnson-Lindenstrauss lemma. Notes on the JL-Lemma |
13. 10/18 | Tue | Midterm Review. | Slides. |
10/20 | Thu | Midterm | Study guide and review questions. |
Spectral Methods | 14. 10/25 | Tue | High-dimensionality geometry and connections to the JL Lemma/dimensionality reduction. | Slides. Compressed slides. Reading: Chapters 2.3-2.6 of Foundations of Data Science on high-dimensional geometry. |
15. 10/27 | Thu | Intro to principal component analysis, low-rank approximation, data-dependent dimensionality reduction. Orthogonal bases and projection matrices. | Slides. Compressed 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 for linear algebra review. Some other good videos. | 16. 11/01 | Tue | Best fit subspaces and optimal low-rank approximation via eigendecomposition. | Slides. Compressed 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. |
17. 11/03 | Thu | Eigenvalues as a measure of low-rank approximation error. The singular value decomposition and connections to low-rank approximation. | Slides. Compressed slides. Reading: Notes on SVD and its connection to eigendecomposition/PCA (Roughgarden and Valiant at Stanford). Chapter 3 of Foundations of Data Science and Chapter 11 of Mining of Massive Datasets on low-rank approximation and the SVD. |
18. 11/08 | Tue | Applications of low-rank approximation beyond compression. Matrix completion, entity embeddings, and non-linear dimensionality reduction. | Slides. Compressed slides. Reading: Notes on matrix completion, with proof of recovery under incoherence assumptions (Jelani Nelson at Harvard). Levy Goldberg paper on word embeddings as implicit low-rank approximation. |
19. 11/10 | Thu | Spectral graph theory and spectral clustering. | Slides. Compressed 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). |
20. 11/15 | Tue | The stochastic block model. | Slides. Compressed 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). |
21. 11/17 | Thu | Computing the SVD: power method. Krylov methods. Connection to random walks and Markov chains. | Slides. Compressed 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). |
11/22 | Tue | No Class. Friday class schedule followed. | |
11/24 | Thu | No Class. Thanksgiving recess. | |
Optimization | |||
22. 11/29 | Tue | Class Canceled | |
23. 12/01 | Thu | Start on optimization and gradient descent. | Slides. Compressed slides. Reading: Chapters I and III of these notes (Hardt at Berkeley). Multivariable calc review, e.g., through: Khan academy |
24. 12/06 | Tue | Gradient descent analysis for convex Lipschitz functions. | Slides. Compressed slides. Reading: Chapters I and III of these notes (Hardt at Berkeley). |
25. 12/08 | Thu | Constrained optimization and projected gradient descent. Course conclusion/review. | Slides. Compressed slides. |
12/14, 10:30am - 12:30pm | Final Exam. | Study guide and review questions. See Moodle for practice exams. |