Tue, Nov 29
We meet at 10A as usual (i.e., 10:00am to 11:50am), in Sudikoff 115.
CR-Precis: A Deterministic Summary Structure for Update Data Streams.
[Yilong + Zhipeng]
Estimating Rarity and Similarity over Data Stream Windows.
[Chen + Vijay]
Estimating the Sortedness of a Data Stream.
[Zhiyu + Zhong]
Space-Efficient Online Computation of Quantile Summaries.
[Shahrzad + Shrirang]
Mon, Dec 5
We meet during this course's final exam slot (i.e., 3:00pm to 5:00pm), in Sudikoff 115.
The Shifting Sands Algorithm.
[Thomas + Yu-Han]
Recognizing Well-Parenthesized Expressions in the Streaming Model.
[Andrew + Huiting]
Estimating PageRank on Graph Streams.
[Cole + Fabio]
- Each presentation should be for 20 minutes.
- Divide your time so that each person speaks for 10 minutes.
- Use good slides, with helpful pictures. Don't just put lots of
text and formulas on the slides.
- Show at least one algorithm in pseudocode form. Don't just lift
pseudocode right out of the paper unless it's very clear pseudocode.
Instead, simplify the paper's presentation (possibly losing some
accuracy and details).
- As a rule of thumb, a 20-minute talk should have 10 to 12 slides.
- Practise your talk to ensure that you can finish within 20
minutes. This is very important!