The StarCraft Multi-Agent Challenges+ : Learning of Multi-Stage Tasks and Environmental Factors without Precise Reward Functions

5 Jul 2022  ·  Mingyu Kim, Jihwan Oh, Yongsik Lee, Joonkee Kim, SeongHwan Kim, Song Chong, Se-Young Yun ·

In this paper, we propose a novel benchmark called the StarCraft Multi-Agent Challenges+, where agents learn to perform multi-stage tasks and to use environmental factors without precise reward functions. The previous challenges (SMAC) recognized as a standard benchmark of Multi-Agent Reinforcement Learning are mainly concerned with ensuring that all agents cooperatively eliminate approaching adversaries only through fine manipulation with obvious reward functions. This challenge, on the other hand, is interested in the exploration capability of MARL algorithms to efficiently learn implicit multi-stage tasks and environmental factors as well as micro-control. This study covers both offensive and defensive scenarios. In the offensive scenarios, agents must learn to first find opponents and then eliminate them. The defensive scenarios require agents to use topographic features. For example, agents need to position themselves behind protective structures to make it harder for enemies to attack. We investigate MARL algorithms under SMAC+ and observe that recent approaches work well in similar settings to the previous challenges, but misbehave in offensive scenarios. Additionally, we observe that an enhanced exploration approach has a positive effect on performance but is not able to completely solve all scenarios. This study proposes new directions for future research.

PDF Abstract

Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
SMAC+ Def_Armored_parallel IQL Median Win Rate 0.0 # 6
SMAC+ Def_Armored_sequential IQL Median Win Rate 9.4 # 8
SMAC+ Def_Infantry_parallel IQL Median Win Rate 40.0 # 8
SMAC+ Def_Infantry_sequential IQL Median Win Rate 93.8 # 7
SMAC+ Def_Outnumbered_parallel IQL Median Win Rate 0.0 # 4
SMAC+ Def_Outnumbered_sequential IQL Median Win Rate 0.0 # 5
SMAC+ Off_Hard_parallel IQL Median Win Rate 0.0 # 3
SMAC+ Off_Superhard_parallel IQL Median Win Rate 0.0 # 1

Methods


No methods listed for this paper. Add relevant methods here