3 ms·
Train a ML system on a range of parameters (whirlpool strength) until you have a decent port of the game to a neural network and/or tree-based algo. Then try to
by thanatropism 7y ago
Train a ML system on a range of parameters (whirlpool strength) until you have a decent port of the game to a neural network and/or tree-based algo. Then try to optimize the game based on people’s enjoyment.
- im3w1l 7y agoI think it would be much easier to use change the parameters directly in the original program. Just have to use a gradient-free optimizer.
- thanatropism 7y agoYours is a better technical approach to the vague dream I had described. Am I being downvoted because I don't have an optimal solution to some problem?