Course Announcement
Many important algorithms and theorems of theoretical
computer science get their power from extremal combinatorics.
This course will cover material from extremal combinatorics that
I feel is useful or could be useful for computer science.
The course is designed for both combinatorics and computer science
students.
Combinatorics students will learn how combinatorics can be applied to
problems of computer science, and computer science students will
encounter many tools (especially orthogonal polynomials) that are
standard in combinatorics, but have not yet made their way into
theoretical computer science.
Everyone involved should expect to encounter interesting
research problems as well as learn techniques that they will find
useful elsewhere in their research.
As such a course has not been taught before at MIT, we will produce
lecture notes.
The following is a list of the topics that I presently plan to discuss.
There are many relations among the topics listed, but I have
forced them into a tree for ease of viewing.
- Eigenvalues of graphs
- Second eigenvalue/mixing rate
- Isoperimetric number, multicommodity flow, geometric
embeddings and isoperimetry
- Spectral partitioning,
geometric embeddings and eigenvalues
- Mixing rates of walks, estimation of volume of
convex bodies
- Isomorphism testing
- Cospectral graphs
- Relation of eigenvalues to automorphisms and
testing isomorphism of graphs of bounded eigenvalue
multiplicity
- Pseudo-random dense graphs
- Discrepancy bounds,
- Paley graphs and Weil sums
- A little Ramsey theory
- Expander graphs
- Random expanders
- Explicit construction (probably Gabber-Galil)
- Application to re-using random bits
- Miscellaneous cute facts (if time allows such frivolity)
- Eigenvectors of graphs
- Planar graphs
- Regular polytopes
- Cayley graphs
- Strongly-regular graphs
- Relation to spherical designs, extremal point sets, and
density of sphere packings
- Limits of regularity
- Association schemes, Bose-Mesner algebra, and
design of statistical experiments
- Error-correcting codes
- Lower bounds
- Naive volume bound
- Elias's bound
- Linear Programming bound
- Constructions
- Gilbert-Varshamov bound (random constructions)
- Reed-Solomon Codes
- Justesen Codes
- Codes from expander graphs and superconcentrators
- Weight distributions of codes
- Constructing expanders from error-correcting codes
- A little extremal set theory
- Erdos-Ko-Rado theorem
- Ray-Chaudhuri--Wilson theorem
- Hashing
(material might be pushed into next semester--advanced
complexity theory)
- Public coins versus private coins
- Hash functions from error-correcting codes
Daniel A. Spielman
Last modified: Mon Sep 9 11:40:19 1996