Heuristic Search
摘要
To this point in Part II we have introduced graph theory, ► Chap. 3, and several algorithms for searching graphs, ► Chap. 4. Our search algorithms to this point have been “blind” with no particular knowledge about the searched problem. ► Chap. 5, on heuristic search, addresses this issue. ► Chapter 5 has four parts. The first is an introduction to heuristics and how these can be used to find “better” paths through a search space. ► Sect. 5.2 presents hill-climbing, a simple greedy search algorithm that takes the “best” choice at every opportunity. We will demonstrate how hill-climbing sometimes reaches suboptimal conclusions. Next, we introduce the best-first search algorithm that can overcome the limitations of hill-climbing search. Finally, ► Sect. 5.4 presents genetic algorithms, an example of exploring heuristic search along parallel paths.