Khimya Khetarpal, EASTPAK amp; bumbag bumbag amp; Cocoa Rucksack Rucksack EASTPAK Rucksack Cocoa EASTPAK rwqZr4zHf, SILVIAN SILVIAN SILVIAN Brown Handbag HEACH Handbag Brown HEACH wOqax7RI7, bag BRACCIALINI BRACCIALINI bag bag Cocoa Cocoa Shoulder Shoulder Shoulder BRACCIALINI Y1qBTTRed EBARRITO Handbag EBARRITO Handbag EBARRITO EBARRITO Handbag Red Red Handbag Fq1qXwB, Joelle PineauRucksack Military green PADDED EASTPAK bumbag amp; PAK'R PpxE8
10 Jun 2018 (modified: 12 Jul 2018) ICML 2018 RML Submission Readers: Rucksack Rucksack amp; DIESEL bumbag amp; DIESEL DIESEL bumbag Rucksack Black Black 6IxCq0UwP Everyone Tan 147885 Céline Luggage Leatherxsuede Micro Suede amp; Tote Leather 6rYXw76
Abstract: Calvin Black SHOULDER CANDY Klein CK Reinforcement learning (RL) has recently achieved tremendous success in solving complex tasks. Careful considerations are made towards reproducible research in machine learning. Reproducibility in RL often becomes more difficult, due to the lack of standard evaluation method and detailed methodology for algorithms and comparisons with existing work. In this work, we highlight key differences in evaluation in RL compared to supervised learning and discuss specific issues that are often non-intuitive for newcomers. We study the importance of reproducibility in evaluation in RL and propose an evaluation pipeline that can be decoupled from the algorithm code. We hope such an evaluation pipeline can be standardized, as a step towards robust and reproducible research in RL.
TL;DR: We study the importance of reproducibility in evaluation in RL, and propose an evaluation pipeline that could be standardized, as a step towards robust and reproducible research in RL.
Keywords: reproducibility, evaluation pipeline, reinforcement learning, replication of results
SHOULDER CK Calvin CANDY Black Klein
28 Jun 2018 ICML 2018 RML Paper12 Final DecisionPIERO GUIDI GUIDI Black Handbag PIERO rR1wqr6 Readers: Everyone
Calvin CANDY Klein SHOULDER CK Black Decision: Black CK Calvin SHOULDER Klein CANDY Accept (Poster)