Atari Games On Atari 2600 Bowling

评估指标

Score

评测结果

各个模型在此基准测试上的表现结果

模型名称	Score	Paper Title	Repository
Advantage Learning	57.41	Increasing the Action Gap: New Operators for Reinforcement Learning	-
GDI-I3	201.9	GDI: Rethinking What Makes Reinforcement Learning Different From Supervised Learning	-
DDQN (tuned) hs	69.6	Deep Reinforcement Learning with Double Q-learning	-
DNA	181	DNA: Proximal Policy Optimization with a Dual Network Architecture	-
A3C LSTM hs	41.8	Asynchronous Methods for Deep Reinforcement Learning	-
CGP	85.8	Evolving simple programs for playing Atari games	-
GDI-H3	205.2	Generalized Data Distribution Iteration	-
IMPALA (deep)	59.92	IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures	-
Duel noop	65.5	Dueling Network Architectures for Deep Reinforcement Learning	-
ASL DDQN	62.4	Train a Real-world Local Path Planner in One Hour via Partially Decoupled Reinforcement Learning and Vectorized Diversity	-
DQN noop	50.4	Deep Reinforcement Learning with Double Q-learning	-
Ape-X	17.6	Distributed Prioritized Experience Replay	-
IQN	86.5	Implicit Quantile Networks for Distributional Reinforcement Learning	-
Duel hs	65.7	Dueling Network Architectures for Deep Reinforcement Learning	-
QR-DQN-1	77.2	Distributional Reinforcement Learning with Quantile Regression	-
A3C FF hs	35.1	Asynchronous Methods for Deep Reinforcement Learning	-
DDQN (tuned) noop	68.1	Dueling Network Architectures for Deep Reinforcement Learning	-
RUDDER	179	RUDDER: Return Decomposition for Delayed Rewards	-
Persistent AL	71.59	Increasing the Action Gap: New Operators for Reinforcement Learning	-
Gorila	54	Massively Parallel Methods for Deep Reinforcement Learning	-

0 of 44 row(s) selected.