George Mason University
Statistics Colloquium Series
Seminar Announcement


Kernel-Based Density Estimators, Local Polynomial Regression, and Particle Methods II: Applications and Preliminary Results

J. Wallin and A. Haque


George Mason University


ABSTRACT

As discussed in the part I, the numerical methods used to solve PDE's and ODE's have considerable difficulties dealing with highly distorted grids. In this talk, we will show some of our work using Moving Least Squares and Local Polynomial Regression to solve systems of ODE and PDE's. Although this work is still in a preliminary stage, we feel it shows excellent promise ad a PDE/ODE solver for systems with highly distorted grids.


Friday, October 6, 2000
George W. Johnson Center, Assembly Room B
Seminar at 10:45 a.m.
Refreshments at 10:30 a.m.
For the 2000 Fall Seminar Schedule, go to
www.science.gmu.edu/statseminars