OpenML Reinforcement Learning Control Benchmark

OpenML's 'support' dataset is a benchmark for reinforcement learning agents, categorized under Agent Training and Control. Platform tags indicate it contains tabular data designed for evaluating control algorithms. Its specific origin, size, and creation date are not documented in the provided metadata.

Use Cases

• Training RL agents on state-action-reward transitions within the tabular control environment.
• Benchmarking policy performance against other algorithms using the provided reward and state features.
• Analyzing agent behavior and learning curves across episodes in a standardized control task.

Strengths

• Part of the OpenML platform, ensuring a standardized format for comparison.
• Explicitly tagged for reinforcement learning and control, providing clear domain relevance.

Limitations

• No information on the number of rows, episodes, or state dimensions is available.
• The absence of column names or sample data prevents assessment of feature engineering or data complexity.

Provenance

Source

OpenML platform