George Mason University
CSI/Statistics Colloquium Series
Seminar Announcement

---

On the Inference of Semi-Coherent Structures from Data

Dean H. Judson


U.S. Bureau of the Census


ABSTRACT This presentation examines a mathematical structure known under various names: It is known as a coherent structure, a monotonic boolean function, a Sperner set, a clutter, a weighted voting rule, and a simple game. We first focus on describing the different contexts in which this structure has been studied, followed by a voting rule interpretation of the structure and an examination of the lattice properties of the structure. Our main focus in this presentation, however, is the inductive inference problem, that is, how to learn a semicoherent structure from data which partially define the structure. We develop a set of algorithms and simulate their success in learning an arbitrary structure, and conclude that the algorithms are feasible.

Friday, December 3, 1999
George W. Johnson Center, Assembly Room B
Seminar at 10:45 a.m.
Refreshments at 10:30 a.m.
For the 1999 Fall Seminar Schedule, go to
http:www.science.gmu.edu/statseminars