The Multi-Agent Reinforcement Learning in MalmÖ (MARLÖ) Competition

• DOI: 10.48550/arxiv.1901.08129
• Publication Date: 2019-01-01
• Authors: Diego Pérez-Liébana; Katja Hofmann; Sharada P. Mohanty; Noburu Kuno; André Krämer; Sam Devlin; Raluca D. Gaina; Daniel Ionita
• Keywords: CONTEST; Milestone; Sophistication

Learning in multi-agent scenarios is a fruitful research direction, but current approaches still show scalability problems multiple games with general reward settings and different opponent types. The Multi-Agent Reinforcement MalmÖ (MARLÖ) competition new challenge that proposes this domain using 3D games. goal of contest to foster agents can learn across types, proposing as milestone the direction Artificial General Intelligence.