Minimax Weight and Q-Function Learning for Off-Policy Evaluation

Authors: Masatoshi Uehara, Jiawei Huang, Nan Jiang

ICML 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We empirically demonstrate the effectiveness of our methods and compare them to baseline algorithms in Cart Pole with function approximation. We compare MWL & MQL to MSWL (Liu et al., 2018, with estimated behavior policy) and Dual DICE (Nachum et al., 2019a). We use neural networks with 2 hidden layers as function approximators for the main function classes for all methods, and use an RBF kernel for the discriminator classes (except for Dual DICE); due to space limit we defer the detailed settings to Appendix E. Figure 1 shows the log MSE of relative errors of different methods, where MQL appears to the best among all methods.
Researcher Affiliation Academia 1Harvard University, Massachusetts , Boston, USA 2University of Illinois at Urbana-Champaign, Champaign, Illinois, USA.
Pseudocode No The paper does not contain structured pseudocode or algorithm blocks (e.g., clearly labeled algorithm sections or code-like formatted procedures).
Open Source Code No The paper does not provide concrete access to source code for the methodology described, such as a specific repository link or an explicit code release statement.
Open Datasets Yes We empirically demonstrate the effectiveness of our methods and compare them to baseline algorithms in Cart Pole with function approximation. (...) To back up this theoretical finding, we also conduct experiments in the Taxi environment (Dietterich, 2000) following Liu et al. (2018, Section 5)
Dataset Splits No The paper does not provide specific dataset split information (exact percentages, sample counts, citations to predefined splits, or detailed splitting methodology) needed to reproduce the data partitioning.
Hardware Specification No The paper does not provide specific hardware details (exact GPU/CPU models, processor types with speeds, memory amounts, or detailed computer specifications) used for running its experiments.
Software Dependencies No The paper mentions using "neural networks" and "RBF kernel" but does not provide specific ancillary software details, such as library or solver names with version numbers.
Experiment Setup No We use neural networks with 2 hidden layers as function approximators for the main function classes for all methods, and use an RBF kernel for the discriminator classes (except for Dual DICE); due to space limit we defer the detailed settings to Appendix E.