Clustering players' strategies for generation of non-player characters | Rok Piltaver | DSC Europe23 

Developing high-quality non-player characters (NPCs) was essential because it could make or break a game. However, it required a lot of game design, development, and quality assurance effort. In the talk, we demonstrated how we applied machine learning to speed up the process, specifically using clustering of players' army formations to generate human-like NPCs in Mythic Legends - an auto-chess inspired strategy role-playing mobile game. We explained why defining a good distance function needed for clustering was almost impossible and how we pragmatically simplified the original clustering problem into a series of clustering steps with partial distance functions, and how we tamed the combinatorial explosion with a greedy algorithm.
This speech by Rok Piltaver was held on November 22nd at Data Science Conference Europe 2023 in Belgrade.
Follow us on social media :
LinkedIn: / admin
Instagram: / datasciconf
Facebook page: / datasciconference
Website: datasciconfere...
Change the world through data
#datascience #ai #ml #artificialintelligence #bigdata #dataanalytics #business

Наука