[Previous] | [Session 91] | [Next]

[91.04U] Genetic Algorithms and Neural Networks Applied to Galaxy Collisions

G. Edirisinghe (Dept. Physics and Astronomy, U. Tennessee; Physics Dept., Johns Hopkins U.), D. Edirisinghe (Dept. Physics and Astronomy, U. Tennessee; Comp. Sci. Dept, Carnegie Mellon U.), O. Messer, M. Guidry (Dept. Physics and Astronomy, U. Tennessee)

We report on progress in the development of general-purpose algorithms for global parameter minimization in scientific applications where comparison of results with data takes the form of a pattern recognition problem. Our basic approach is to implement model calculations for the problem of interest in parallel on a Beowulf cluster, with a genetic algorithm to optimize the parameters of the calculation and a neural network trained on observational data to compute the fitness function for members of the genetic population in each successive generation. In the specific application discussed in this presentation, we investigate galaxy collisions using a gravity tree plus SPH hydrodynamics (the Max Planck code GADGET), our own genetic algorithm code for global parameter minimization, and our own neural network code for comparison of calculations with observational data.

[Previous] | [Session 91] | [Next]

Bulletin of the American Astronomical Society, 34, #4
© 2002. The American Astronomical Soceity.