Learning Action Representations for Reinforcement Learning

Authors: Yash Chandak, Georgios Theocharous, James Kostas, Scott Jordan, Philip Thomas

ICML 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental 5. Empirical Analysis To understand the internal working of our proposed algorithm, we present visualizations of the learned action representations on the maze domain.
Researcher Affiliation Collaboration 1University of Massachusetts, Amherst, USA. 2Adobe Research, San Jose, USA.
Pseudocode Yes Algorithm 1: Policy Gradient with Representations for Action (PG-RA)
Open Source Code No The paper does not provide an explicit statement or link for open-source code release.
Open Datasets No For both of these applications, an existing log of user s click stream data was used to create an n-gram based MDP model for user behavior (Shani et al., 2005). In the tutorial recommendation task, user activity for a three month period was observed. Sequences of user interaction were aggregated to obtain over 29 million clicks. Similarly, for a month long duration, sequential usage patterns of the tools in the multi-media editing software were collected to obtain a total of over 1.75 billion user clicks.
Dataset Splits No The paper describes the data sources but does not provide specific details on training, validation, or test dataset splits, percentages, or methodology for splitting.
Hardware Specification No The paper does not provide specific details on the hardware used for running experiments (e.g., CPU, GPU models, or memory).
Software Dependencies No The paper does not list specific software dependencies with version numbers (e.g., Python, PyTorch, TensorFlow versions or other libraries).
Experiment Setup Yes For detailed discussion on parameterization of the function approximators and hyper-parameter search, see Appendix D.