Reinforcement Learning for Control with Multiple Frequencies
Authors: Jongmin Lee, Byung-Jun Lee, Kee-Eung Kim
NeurIPS 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We conducted a set of experiments in order to evaluate the effectiveness of AP-AC on high-dimensional tasks with different control frequencies. To the best of our knowledge, this work is the first to address multiple control frequencies in RL. Since there are no existing RL methods designed for multiple control frequencies, we take the variants of SAC as baselines for performance comparison, which are listed as follows: |
| Researcher Affiliation | Academia | Jongmin Lee1, Byung-Jun Lee1, Kee-Eung Kim1,2 1 School of Computing, KAIST, Republic of Korea 2 Graduate School of AI, KAIST, Republic of Korea {jmlee,bjlee}@ai.kaist.ac.kr, kekim@kaist.ac.kr |
| Pseudocode | Yes | The pseudo-code of AP-PI can be found in Appendix D. The pseudo-code for AP-AC can be found in Appendix E. |
| Open Source Code | No | The paper does not provide an explicit statement about releasing its source code for the proposed methods (AP-PI or AP-AC), nor does it include a link to a code repository. It mentions external tools like SUMO-RL [1] and OpenAI Gym [3] but not its own implementation code. |
| Open Datasets | Yes | We conducted experiments on four Open AI Gym continuous control tasks based on the Mujoco physics simulator [3, 26]... We use SUMO (Simulation of Urban MObility) [8] as the traffic simulator and SUMO-RL [1] for the environment interface. |
| Dataset Splits | No | The paper does not provide specific train/validation/test dataset splits with percentages or sample counts. As a reinforcement learning paper, it describes training agents in continuous control tasks and a traffic simulation environment, where data is generated through interaction rather than from pre-defined static splits. |
| Hardware Specification | No | The paper does not provide specific details about the hardware used for running experiments, such as GPU models, CPU specifications, or memory. |
| Software Dependencies | No | The paper mentions using 'SUMO (Simulation of Urban MObility) [8] as the traffic simulator and SUMO-RL [1] for the environment interface.' However, it does not specify version numbers for these or any other software dependencies, which would be necessary for reproducibility. |
| Experiment Setup | Yes | The experimental setups including hyperparameters can be found in Appendix G. |