End-to-End Game-Focused Learning of Adversary Behavior in Security Games
Authors: Andrew Perrault, Bryan Wilder, Eric Ewing, Aditya Mate, Bistra Dilkina, Milind Tambe1378-1386
AAAI 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We test our approach on a combination of synthetic and human subject data and show that game-focused learning outperforms a two-stage approach in settings where the amount of data available is small and when there is wide variation in the adversary s values for the targets. |
| Researcher Affiliation | Academia | 1Center for Research on Computation and Society, Harvard 2Center for Artificial Intelligence in Society, University of Southern California |
| Pseudocode | No | The paper describes its approach and flow with diagrams but does not include any pseudocode or algorithm blocks. |
| Open Source Code | No | The paper does not contain any statement about releasing source code or providing a link to it. |
| Open Datasets | Yes | We use data from human subject experiments performed by Nguyen et al. (2013). |
| Dataset Splits | Yes | Game-tuned two-stage (2S-GT) is a regularized approach that aims to maximize the defender s expected utility when the amount of data is small. It uses Dropout (Srivastava et al. 2014) and a validation set for early stopping. |
| Hardware Specification | No | The paper does not specify any hardware details such as GPU/CPU models, memory, or specific computing environments used for experiments. |
| Software Dependencies | No | The paper mentions implementing neural networks and using gradient descent, but does not provide specific software dependencies or version numbers (e.g., Python, TensorFlow, PyTorch versions). |
| Experiment Setup | Yes | Unless it is varied in an experiment, the parameters are: 1. Number of targets = |T | {8, 24}. 2. Features per target = |y|/|T | = 100. 3. Number of training games = |Dtrain| = 50. ... 6. We fix the attacker s weight on defender coverage to be w = 4 (see Eq. 2)... |