Support2 Control Systems Benchmark for Reinforcement Learning

Support2 is a reinforcement learning dataset from the OpenML platform, categorized for control systems tasks. The dataset is tagged as a benchmark for evaluating agent performance. Specific details on row count, authorship, and temporal coverage are not provided in the available metadata.

Use Cases

• Benchmarking RL agent performance on control system tasks using state and action spaces defined in the dataset.
• Training policy networks on the provided state observations and reward signals.
• Comparing algorithm sample efficiency and final performance against established benchmarks for this environment.

Strengths

• Dataset is hosted on OpenML, a platform with standardized access and versioning.
• Categorized as a benchmark, implying a defined evaluation protocol.

Limitations

• No metadata on dataset size, feature count, or temporal coverage is available.
• Unknown data collection methodology and potential for simulated environment biases.

Provenance

Source

OpenML platform