ETRI-Knowledge Sharing Plaform

ENGLISH

성과물

논문 검색
구분 SCI
연도 ~ 키워드

상세정보

학술대회 An Analysis of the Effects of Physical Abilities on RTS Game Outcome Using Machine Learning Approach
Cited 0 time in scopus Download 1 time Share share facebook twitter linkedin kakaostory
저자
김대욱, 박성윤, 양성일
발행일
201910
출처
International Conference on Information and Communication Technology Convergence (ICTC) 2019, pp.929-933
DOI
https://dx.doi.org/10.1109/ICTC46691.2019.8939771
협약과제
19CS1700, 메타 플레이 인식 기반 지능형 게임 서비스 플랫폼 개발, 양성일
초록
Physical ability in games represents how many actions or multitasks a gamer can do or how fast the gamer can react. In this paper, we characterize physical ability features from Starcraft II data. The proposed win prediction model based on machine learning algorithms evaluates how much features affect a game outcome. Even if features are used only related to physical ability, prediction result shows higher accuracy on a small scale than one with score-related features. Finally, we analyze those features and conclude that although matchmaking rating(MMR) plays primary role to lead a game win, actions per minute and others are also crucial factor to win.
키워드
game analytics, machine learning, prediction, real-time strategy games, starcraft
KSP 제안 키워드
Game Analytics, Machine Learning Algorithms, Machine Learning Approach, RTS games, Real-time strategy games, Small-scale, StarCraft II, Win prediction, model-based, prediction model