Goal Recognition Design for Non-Optimal Agents
Authors: Sarah Keren, Avigdor Gal, Erez Karpas
AAAI 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | In this work we present goal recognition design for non-optimal agents, which extends previous work by accounting for agents that behave non-optimally either intentionally or na ıvely. The analysis we present includes a new generalized model for goal recognition design and the worst case distinctiveness (wcd) measure. For two special cases of sub-optimal agents we present methods for calculating the wcd, part of which are based on novel compilations to classical planning problems. Our empirical evaluation shows the proposed solutions to be effective in computing and optimizing the wcd. |
| Researcher Affiliation | Academia | Sarah Keren and Avigdor Gal {sarahn@tx,avigal@ie}.technion.ac.il Technion Israel Institute of Technology Erez Karpas karpase@csail.mit.edu Massachusetts Institute of Technology |
| Pseudocode | No | The paper does not contain structured pseudocode or algorithm blocks. |
| Open Source Code | No | The paper does not provide concrete access to source code for the methodology described. |
| Open Datasets | Yes | We use the domains proposed by Ramirez and Geffner (2009) for plan recognition. The dataset consists of problems from 4 domains, namely GRID-NAVIGATION, IPC-GRID+, BLOCK-WORDS and LOGISTICS. |
| 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 into training, validation, and test sets. |
| 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 does not provide specific ancillary software details (e.g., library or solver names with version numbers) needed to replicate the experiment. |
| Experiment Setup | Yes | For the Bounded Deception setting we compared the former 4 methods and examined problems with the budget of the agents aiming at POI ranging from 1 to 7 for the GRID-NAVIGATION , IPCGRID+ and BLOCK-WORDS domain and from 1 to 3 for the LOGISTICS (which is the maximal budget the planner could handle). For the Bounded Non-Optimal setting we tested the timed and sync methods with the diversion budget of both agents ranging as for the Bounded Deception setting. |