FANDOM


2019-05-18

The modern literature about Artificial Intelligence is devoted to advanced planning algorithms. A given model is used by a solver to predict the outcome of a system and the induction algorithm is able to create new models from scratch. But before such systems can be introduced, it's important to become familiar with the basics. A simple handcrafted method for creating an ingame AI is called a behavior tree and was described in the tutorial in detail.[1] The possible actions of an agent are ordered hierarchical and it's possible to change the overall structure on demand. A normal behavior tree is less powerful then the GOAP concept, but it make sense to read the tutorial to get an idea how the behavior is stored in a machine readable format in general . Unfortunately, the video which is given in the article is not very communicative and doesn't provide extra information.

[1] https://www.gamasutra.com/blogs/ChrisSimpson/20140717/221339/Behavior_trees_for_AI_How_they_work.php

Community content is available under CC-BY-SA unless otherwise noted.