Tilted Forum Project Discussion Community - View Single Post - Databases, Optimisation and Search Algorithms

manalone · 03-20-2004, 02:34 PM

optimality cannot be acheived in a guaranteed fashion through stochastic methods.

IDA* is a good idea, kel's answer is pretty comprehensive. But I find it hard to think of a goal, unless you can test for optimality arbitrarily.

So you might be looking at B&B.

It's difficult to suggest something without knowing more about the problem, but you might be able to pre-calculate some predictor.

Try examining statistical methods before you do anything else.

Ideally, you could cook up some correlation factor and just extract the best one of those. If you are going down the annealing / genetic line, you must have a fitness function. Why not just apply the fitness function to the data? It means a comprehensive search, but you might be able to subdivide the data by sorting it and bounding it.

03-20-2004, 02:34 PM	#3 (permalink)
manalone Once upon a time...	optimality cannot be acheived in a guaranteed fashion through stochastic methods. IDA* is a good idea, kel's answer is pretty comprehensive. But I find it hard to think of a goal, unless you can test for optimality arbitrarily. So you might be looking at B&B. It's difficult to suggest something without knowing more about the problem, but you might be able to pre-calculate some predictor. Try examining statistical methods before you do anything else. Ideally, you could cook up some correlation factor and just extract the best one of those. If you are going down the annealing / genetic line, you must have a fitness function. Why not just apply the fitness function to the data? It means a comprehensive search, but you might be able to subdivide the data by sorting it and bounding it. __________________ -- Man Alone ======= Abstainer: a weak person who yields to the temptation of denying himself a pleasure. Ambrose Bierce, The Devil's Dictionary. Last edited by manalone; 03-20-2004 at 02:37 PM..