Hi. I'm looking for an efficient implementation of a CP-based shortest-path shortest path algorithm in Eclipse. Does anyone have something like that? I'm writing an AI planner in Eclipse that performs a search in the space of `partial plans', computing a heuristic value at every node in a way that's similar to a Bellman-Ford shortest path algorithm. The code does not run very fast and most of the time seems to be taken by this (repeated) computation. Yet I don't see how to use the information provided by the profiler to make this key routine much faster, and wanted to take a look at a good implemented version of a CP version of Bellman-Ford to see if I'm missing something. I remember having read a paper by Claude Le Pape and others about a CP-based shortest-path algorithm approaching the performance of Bellman-Ford by scheduling the constraints in a suitable way. Any pointer or relevant code would be highly appreciated. Also if there is a serious Eclipse programmer out there who is interested in what's now called heuristic search planning in AI and is willing to help, please let me know. Thanks a lot. Cheers. - Hector Geffner www.ldc.usb.ve/~hectorReceived on Thu Jan 04 23:01:00 2001
This archive was generated by hypermail 2.1.8 : Wed 16 Nov 2005 06:07:07 PM GMT GMT