Constrained Risk-Averse Markov Decision Processes

Authors: Mohamadreza Ahmadi, Ugo Rosolia, Michel D. Ingham, Richard M. Murray, Aaron D. Ames11718-11725

AAAI 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Numerical Experiments In this section, we evaluate the proposed methodology with a numerical example. In addition to the traditional total expectation, we consider two other coherent risk measures, namely, CVa R and EVa R. All experiments were carried out using a Mac Book Pro with 2.8 GHz Quad-Core Intel Core i5 and 16 GB of RAM. The resultant linear programs and DCPs were solved using CVXPY (Diamond and Boyd 2016) with DCCP (Shen et al. 2016) add-on in Python.
Researcher Affiliation Collaboration 1California Institute of Technology, 1200 E. California Blvd., Pasadena, CA 91125 2NASA Jet Propulsion Laboratory, 4800 Oak Grove Dr, Pasadena, CA 91109.
Pseudocode No The paper does not contain any structured pseudocode or algorithm blocks.
Open Source Code No The paper does not provide any statement or link indicating that the authors' implementation code is open-source or publicly available.
Open Datasets No The paper describes a simulated 'grid world' for its experiments, but it does not specify or provide access information for a publicly available or open dataset.
Dataset Splits No The paper describes numerical experiments and Monte Carlo simulations but does not specify any train/validation/test dataset splits.
Hardware Specification Yes All experiments were carried out using a Mac Book Pro with 2.8 GHz Quad-Core Intel Core i5 and 16 GB of RAM.
Software Dependencies No The paper mentions using 'CVXPY' and 'DCCP add-on in Python' but does not provide specific version numbers for these software components.
Experiment Setup Yes The action set available to the robot is Act = {E, W, N, S, NE, NW, SE, SW}, i.e., diagonal moves are allowed. ... The discount factor is γ = 0.95. ... we set ε = 0.15 for CVa R and EVa R coherent risk measures. The fuel budget (constraint bound β) was set to 50, 10, and 200 for the 10 10, 15 15, and 20 20 grid-worlds, respectively.