课程: AI Algorithms for Game Design with Python

免费学习该课程！

今天就开通帐号，24,700 门业界名师课程任您挑！

Writing good evaluation functions

Writing good evaluation functions - Python教程

课程: AI Algorithms for Game Design with Python

开始一个月试用 购买团队方案

Writing good evaluation functions

“

- [Instructor] So here's what depth limited search does. We select a constant level at which we wish to stop the search, and this will leave us with a very valuable smaller tree. Notice that we will not analyze most of the original tree, and that's fine. The running time for the algorithm on this tree depends on the level at which we are willing to stop. So some trial and error may help us decide this level. However, this improvement in time comes with a cost, and this cost is accuracy. At the bottom of this smaller tree, we need a good evaluation function to estimate how well we'll do if we take the paths underneath these bottom nodes. If we use a poor evaluation function, our agent will behave poorly. Remember that Minimax produces the best solution possible, so that's the best we could aim for, and we must be ready to have a less than perfect agent. In order to minimize these imperfections, we need to come up with a good evaluation function. Here we have three important details to…

内容