Evolutionary Diversity Optimization with Clustering-based Selection for Reinforcement Learning

Authors: Yutong Wang, Ke Xue, Chao Qian

ICLR 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments on various (i.e., deceptive and multi-modal) continuous control tasks, show the superior performance of EDO-CS over previous methods, i.e., EDO-CS can achieve a set of policies with both high quality and diversity efficiently while previous methods cannot.
Researcher Affiliation Academia Yutong Wang , Ke Xue and Chao Qian State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing 210023, China {wangyt, xuek, qianc}@lamda.nju.edu.cn
Pseudocode Yes Algorithm 1: EDO-CS
Open Source Code No The paper does not provide an explicit statement or link to its open-source code.
Open Datasets Yes To examine the performance of EDO-CS, we conduct experiments on a variety of continuous control tasks from Open AI Gym library (Brockman et al., 2016).
Dataset Splits No The paper mentions training and testing environments but does not explicitly describe a validation dataset split or its usage.
Hardware Specification No The paper does not provide specific details about the hardware used for experiments, such as CPU or GPU models.
Software Dependencies No The paper mentions 'Open AI Gym library (Brockman et al., 2016)' and 'Mu Jo Co environments' but does not provide specific version numbers for these or other software dependencies.
Experiment Setup Yes For EDO-CS, the number M of candidates for selection in each cluster is set to 2, and the arms of the bandit are {λ(1) = 0, λ(2) = 0.5}. Other parameter settings can be found in Appendix A.1. (Appendix A.1 provides tables with specific numerical values for Population size M, Archive size l, Number T of updating iterations, σ in ES, and η in ES for different environments.)