6.856J / 18.416J Randomized Algorithms

As taught in: Fall 2002

Unstructured grid for a four element airfoil.

Partitioning algorithms are used to solve complex large scale computational problems, as shown in this unstructured grid for a four element airfoil. (Image is taken from NASA's Web site: http://www.nasa.gov.)

Level:

Graduate

Instructors:

Prof. David R. Karger

Course Features

Course Description

This course examines how randomization can be used to make algorithms simpler and more efficient via random sampling, random selection of witnesses, symmetry breaking, and Markov chains. Topics covered include: randomized computation; data structures (hash tables, skip lists); graph algorithms (minimum spanning trees, shortest paths, minimum cuts); geometric algorithms (convex hulls, linear programming in fixed or arbitrary dimension); approximate counting; parallel algorithms; online algorithms; derandomization techniques; and tools for probabilistic analysis of algorithms.