DQN Boltzmann mountain car benchmark

[kengz]

Experiment Result

See PR #180 for an example

The data files to upload below are all created automatically in the data/ folder.

Github does not support .json and .csv upload, but .txt. Please rename those files .txt before uploading.

Abstract

DQN Boltzmann Mountain Car Benchmark

Methods

No state normalization since env positions are not invariant to translation.

To Reproduce

1. JSON spec:
   dqn_boltzmann_mountain_car_spec.txt

2. git SHA (contained in the file above): 67f44f06a9be5ca420578aefdfb1eb45a291450d

Results

All the results contributed will be added to the benchmark, and made publicly available on Dropbox.

• 1. full experiment data zip: (please find our contact in README and request a "Dropbox file request" to upload it to the public benchmark folder.)

• 2. experiment graph:
  

• 3. max fitness score and experiment_df: 1.04,
  dqn_boltzmann_mountain_car_experiment_df.txt

• 4. best trial JSON spec: dqn_boltzmann_mountain_car_t125_spec.txt

• 5. best trial graph:
  

• 6. [optional] best session graph:
  

Discussion (optional)

Describe some useful observations from the experiment.

Turns our state normalization is harmful in this environment, which makes sense since normalizing would lose information about the starting and end point in the environment, which are absolute. The problem is not invariant to translation and scaling.