Instructors
Edith Cohen, Professor 
Amos Fiat, Professor
Haim Kaplan, ProfessorClass Schedule
| Week | Date | Instructor | Topics | Slides | Notes |
|---|---|---|---|---|---|
| 1 | 24/10/2017 | Edith Cohen |
Distributed and streamed data. Approximate summary structures. Frequent elements: Misra Gries Summaries Set memebership: Bloom filters Counting: Morris approximate counters Distinct counting: preview | PPT PDF | |
| 2 | 31/10/2017 | Edith Cohen |
Sampling-based summaries (MinHash sketches): Distinct counting Reservoir sampling, Distinct sampling Domain-size estimators | PPT PDF | |
| 3 | 7/11/2017 | Edith Cohen |
Sampling-based summaries (continued)
Set relations (union size, Jaccard, cosine similarity) Entities as sets of features: Broder's n-grams for text documents Estimating set relations from MinHash sketches weighted sampling: Poisson PPS, multiobjective Bottom-k samples/sketches Intro to Linear sketches (random projections) |
PPT PDF | |
| 4 | 14/11/2017 | Haim Kaplan |
Linear sketches (random projections) Second frequency moment: The AMS sketch Dimensionality reduction: Johnson-Lindenstrauss transform | PPT PDF | |
| 5 | 21/11/2017 | Haim Kaplan |
Dimensionality reduction (continued) Matrix factorizations: PCA, SVD | notes (PDF) | |
| 6 | 28/11/2017 | Amos Fiat | Data Privacy | PPT PDF | |
| 7 | 5/12/2017 | Uri Stemmer |
Practical local differential privacy Users retain their data and send safe randomizations. | notes (PDF) | Guest lecture |
| 8 | 12/12/2017 | Haim Kaplan | Nearest neighbor search/classifiers Locality Sensitive Hashing (LSH) | PPT PDF | |
| 9 | 19/12/2017 | Edith Cohen |
Graph mining/learning: Introduction Eigenvector centrality: Degree to PageRank MinHash sketches of reachability sets: (reach centrality, similarity, influence) Efficient computation. | PPT PDF | |
| 10 | 26/12/2017 | Edith Cohen |
Graph mining/learning (continued): Distance-based graph mining All-distance sketches (ADS): Efficient computation, Applications | PPT PDF | |
| 11 | 2/1/2018 | Edith Cohen |
Graph mining/learning (continued): ADS application: Approximate distance oracles Submodular set functions: Definition, properties, applications Maximizing submodular monotone functions: (Approximate) Greedy maximization | PPT PDF | tentative |
| 12 | 9/1/2018 | Edith Cohen |
Graph mining/learning (continued) Graph-based influence maximization Semi-supervised learning | PPT PDF | tentative |
| 13 | 16/01/2018 | Haim Kaplan | clustering | PPT PDF |