Loss-Calibrated Monte Carlo Action Selection
Authors: Ehsan Abbasnejad, Justin Domke, Scott Sanner
AAAI 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Empirically, we show that using our loss-calibrated Monte Carlo method yields high-accuracy optimal action selections in a fraction of the number of samples required by conventional loss-insensitive samplers. ... We evaluate our loss-calibrated Monte Carlo method in two domains. We first examine synthetic plant control examples ... We also demonstrate results in a Bayesian decision-theoretic robotics setting with uncertain localization... |
| Researcher Affiliation | Academia | Ehsan Abbasnejad ANU & NICTA ehsan.abbasnejad@anu.edu.au Justin Domke NICTA & ANU justin.domke@nicta.com.au Scott Sanner NICTA & ANU scott.sanner@nicta.com.au |
| Pseudocode | No | The paper does not contain a clearly labeled pseudocode or algorithm block. |
| Open Source Code | No | The paper does not provide any statement or link indicating that the source code for the methodology is openly available. |
| Open Datasets | No | The paper uses synthetic datasets (power plant control and robotics environments with defined utility functions and probability distributions). It does not provide access information (link, DOI, repository, or formal citation with authors/year) for a publicly available or open dataset. |
| Dataset Splits | No | The paper describes simulations and uses synthetic data, but it does not specify explicit training, validation, or test dataset splits in terms of percentages, sample counts, or by referencing predefined splits. |
| Hardware Specification | No | The paper does not provide specific hardware details (e.g., GPU/CPU models, memory, or cloud instance types) used for running its experiments. |
| Software Dependencies | No | The paper mentions using 'Metropolis-Hastings MCMC' but does not specify any software names with version numbers (e.g., programming languages, libraries, or solvers). |
| Experiment Setup | Yes | In each experiment n samples are generated 200 times and the mean of the percentage of times the true optimal action is selected is reported. We use Metropolis-Hastings MCMC by initializing the chain at a random point and using a Normal distribution centered at the current sample with isotropic covariance optimally tuned so that around 23% of samples are accepted (Roberts, Gelman, and Gilks 1997). |