UNIFORMLY DISTRIBUTED DATA EFFECTS IN OFFLINE RL: A CASE STUDY IN GRIDWORLD SETTING
| dc.contributor.author | Tokayev, Kuanysh; Park, Jurn-Gyu | |
| dc.date.accessioned | 2024-05-24T11:12:56Z | |
| dc.date.available | 2024-05-24T11:12:56Z | |
| dc.date.issued | 2024-05-24 | |
| dc.description.abstract | In the emerging landscape of off-policy reinforcement learning (RL), challenges arise due to the significant costs and risks tied to data collection. To address these issues, there is an alternative path for transitioning from off-policy to offline RL, known for its fixed data collection practices. This stands in contrast to online algorithms, which are sensitive to changes in data during the learning phase. However, the inherent challenge of offline RL lies in its limited interaction with the environment, resulting in inadequate data coverage. Hence, we underscore the convenient application of offline RL, 1) starting from thecollection of a static dataset, 2) followed by the training of offline RL models, and 3) culminating with testing in the same environment as off-policy RL methodologies. This involves the utilization of a uniform dataset gathered systematically via non- arbitrary action selection, covering all possible states of the environment. Utilizing the proposed approach, the Offline RL model employing a Multi-Layer Perceptron (MLP) achieves a testing accuracy that falls within 1% of the results obtained by the off-policy RL agent. Moreover, we provide a practical guide with datasets, offering valuable tutorials on the application of Offline RL in Gridworld-based real-world applications. | en_US |
| dc.identifier.citation | Tokayev, Kuanysh, & Park, Jurn-Gyu (2024). Uniformly distributed data effects in offline RL: A case study in Gridworld setting. Nazarbayev University School of Engineering and Digital Sciences. | en_US |
| dc.identifier.uri | http://nur.nu.edu.kz/handle/123456789/7709 | |
| dc.language.iso | en | en_US |
| dc.publisher | Nazarbayev University School of Engineering and Digital Sciences | en_US |
| dc.rights | Attribution 3.0 United States | * |
| dc.rights.uri | http://creativecommons.org/licenses/by/3.0/us/ | * |
| dc.subject | Type of access: Open Access | en_US |
| dc.subject | offline RL | en_US |
| dc.subject | data distribution | en_US |
| dc.subject | deep learning | en_US |
| dc.subject | DQN | en_US |
| dc.subject | machine learning | en_US |
| dc.subject | tutorial | en_US |
| dc.title | UNIFORMLY DISTRIBUTED DATA EFFECTS IN OFFLINE RL: A CASE STUDY IN GRIDWORLD SETTING | en_US |
| dc.type | Technical Report | en_US |
| workflow.import.source | science |