Jan 13 (Sun) |
Jan 14 (Mon) |
Jan 15 (Tue) |
Jan 16 (Wed) |
Jan 17 (Thu) |
Jan 18 (Fri) |
|
10:00am |
Arrival |
Hong Kong Excursion | Individual Meeting with professors & Departure |
|||
12:00noon | Lunch |
|||||
2:00pm |
Registration |
Taking group photo |
||||
Group presentation |
||||||
Tea break |
||||||
4:00pm |
Panel discussion |
|||||
6:00pm |
Closing Dinner |
The theme of the school this year "Computing with noisy data" |
|
January 14, 2013 (Monday) | |
10:00am | Openning |
10:15am - 5:00pm | Invitation to spectral graph theory by Prof. LAU Lap Chi |
Abstract: What is the relation between the eigenvalues of the adjacency matrix of a graph and its combinatorial properties? Surprisingly, eigenvalues and eigenvectors reveal much information about how to partition the graph. This connection is used in designing some of the most efficient heuristics for graph partitioning problems, with applications in different areas of computer science. In two 2-hour sessions, I will introduce the basics of spectral graph theory, prove the Cheeger's inequality relating the second eigenvalue to the graph expansion, and discuss some new developments relating other eigenvalues to graph partitioning. [Lecture Notes] |
|
Problem Session - [Handout] | |
January 15, 2013 (Tuesday) | |
10:00am-5:00pm | Information theory and channel coding theorem by Prof. Chandra Nair |
Abstract: Shannon developed a mathematical theory for communication over noisy medium in the 1940s. One of the main contributions of this theory has been to abstract the key essence of communication and characterize the maximum achievable rate, or channel capacity. In this lecture we will develop the basic tools of information theory and probability needed to establish this classical result. Please click for [Lecture Notes]
|
|
Problem Session - [Handout] | |
January 16, 2013 (Wednesday) | |
10:00am-12:00noon | Using noise to encrypt data by Prof. Andrej Bogdanov |
Abstract: Separating noise from data can be a difficult engineering task. In some cases it is so difficult that we do not know what to do at all. One example is solving linear equations with a noisy right hand side. However it turns out that the hardness of this problem is useful for data encryption: This is a way for Alice and Bob to share secret data over a public communication channel.
I will explain what secure data encryption is and give a blueprint for it under the assumption that it is hard to solve noisy linear equations. |
|