Value-Based Benchmarks

We provide the benchmarks for the following Finite Temporal-Difference algorithms:

• SARSA

• QLearning

• SpeedyQLearning

• WeightedQLearning

• DoubleQLearning

• SARSALambda

• QLambda

We provide the benchmarks for the following Continuous state Temporal-Difference algorithms:

• SARSALambdaContinuous

• TrueOnlineSARSALambda

We provide the benchmarks for the following DQN algorithms:

• DQN

• PrioritiziedDQN

• DoubleDQN

• AveragedDQN

• DuelingDQN

• MaxminDQN

• CategoricalDQN

• NoisyDQN

We consider the following environments in the benchmark

• Finite State Environment Benchmark
  • GridWorld
• Gym Environments Benchmarks
  • MountainCar-v0
• Atari Environment Benchmark
  • BreakoutDeterministic-v4