一句话：提出 MDPFuzz，面向求解 MDP 的 DNN、RL、IL 等模型开展测试，以应对其缺乏充分测试与可靠性保证的问题。
本文关注求解马尔可夫决策过程（MDP）的模型的测试与可靠性问题。MDP 为序贯决策提供数学框架，自主驾驶、机器人控制等安全攸关任务广泛依赖它；深度神经网络、强化学习和模仿学习等 AI 方法可高效求解 MDP，但这些模型既未被充分测试，也缺乏严格可靠性保证。为此，作者提出 MDPFuzz，从标题与摘要看，其目标是对求解 MDP 的模型进行模糊测试或系统化测试，以暴露潜在缺陷。摘要未给出具体测试算法、测试预言、基准与实验结果，因此无法进一步概括其技术细节和定量结论。
The Markov decision process (MDP) provides a mathematical frame- work for modeling sequential decision-making problems, many of which are crucial to security and safety, such as autonomous driving and robot control. The rapid development of artificial intelligence research has created efficient methods for solving MDPs, such as deep neural networks (DNNs), reinforcement learning (RL), and imitation learning (IL). However, these popular models solving MDPs are neither thoroughly tested nor rigorously reliable.
DOI 原文 ·
@inproceedings{paperbot7393,
title = {MDPFuzz: testing models solving Markov decision processes},
author = {Qi Pang and Yuanyuan Yuan and Shuai Wang},
booktitle = {ISSTA},
year = {2022},
doi = {10.1145/3533767.3534388}
}