Fully Parameterized Quantile Function for Distributional Reinforcement Learning

NeurIPS 2019 Derek YangLi ZhaoZichuan LinTao QinJiang BianTieyan Liu

Distributional Reinforcement Learning (RL) differs from traditional RL in that, rather than the expectation of total returns, it estimates distributions and has achieved state-of-the-art performance on Atari Games. The key challenge in practical distributional RL algorithms lies in how to parameterize estimated distributions so as to better approximate the true continuous distribution... (read more)

PDF Abstract

Results from the Paper


TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK RESULT BENCHMARK
Atari Games Atari 2600 Alien FQF Score 16754.6 # 4
Atari Games Atari 2600 Amidar FQF Score 3165.3 # 5
Atari Games Atari 2600 Asterix FQF Score 578388.5 # 3
Atari Games Atari 2600 Asteroids FQF Score 4553.0 # 9
Atari Games Atari 2600 Battle Zone FQF Score 87928.6 # 4
Atari Games Atari 2600 Berzerk FQF Score 12422.2 # 5
Atari Games Atari 2600 Bowling FQF Score 102.3 # 5
Atari Games Atari 2600 Breakout FQF Score 854.2 # 3
Atari Games Atari 2600 Chopper Command FQF Score 876460.0 # 3
Atari Games Atari 2600 Crazy Climber FQF Score 223470.6 # 5
Atari Games Atari 2600 Fishing Derby FQF Score 52.7 # 5
Atari Games Atari 2600 Frostbite FQF Score 16472.9 # 3
Atari Games Atari 2600 Gravitar FQF Score 1406.0 # 10
Atari Games Atari 2600 HERO FQF Score 30926.2 # 10
Atari Games Atari 2600 Ice Hockey FQF Score 17.3 # 5
Atari Games Atari 2600 James Bond FQF Score 87291.7 # 1
Atari Games Atari 2600 Kung-Fu Master FQF Score 111138.5 # 3
Atari Games Atari 2600 Ms. Pacman FQF Score 7631.9 # 5
Atari Games Atari 2600 Phoenix FQF Score 174077.5 # 5
Atari Games Atari 2600 River Raid FQF Score 23560.7 # 5
Atari Games Atari 2600 Robotank FQF Score 75.7 # 3
Atari Games Atari 2600 Skiing FQF Score -9085.3 # 5
Atari Games Atari 2600 Space Invaders FQF Score 46498.3 # 3
Atari Games Atari 2600 Star Gunner FQF Score 131981.2 # 7
Atari Games Atari 2600 Wizard of Wor FQF Score 44782.6 # 5

Methods used in the Paper


METHOD TYPE
🤖 No Methods Found Help the community by adding them if they're not listed; e.g. Deep Residual Learning for Image Recognition uses ResNet