P2 Model-based exploration of the frontier of behaviours for deep learning system testing
面向深度学习系统测试的行为前沿的基于模型探索
FSE Proceedings of the 28th ACM Joint Meeting on European Software Engineering Conference and Symposium on the Foundations of Software Engineering2020匹配度 3/10 · 该论文研究深度学习系统测试，关注模型行为空间与错误区域，属于ML系统可靠性方向；但与用户核心兴趣（面向ML/Agentic程序的静态分析、LLM改进静态分析、浮点程序分析、贝叶斯程序分析）均无直接交集。引用 3inbox
随着深度学习被用于自动驾驶等关键任务，评估依赖 DL 的系统质量变得至关重要。由于训练后的 DL 系统对任意数值向量都会产生输出，无论输入是否位于被测系统有效性域内，其质量取决于有效性域与输出表现为错误行为的区域之间的交集。该论文题为《Model-based exploration of the frontier of behaviours for deep learning system testing》，从标题可见其采用基于模型的方法探索行为前沿，从而面向深度学习系统测试，系统性地寻找并覆盖可能揭示错误行为的前沿区域。给出的摘要仅说明问题设定，未提供具体方法细节、实验设置或结果。
With the increasing adoption of Deep Learning (DL) for critical tasks, such as autonomous driving, the evaluation of the quality of systems that rely on DL has become crucial. Once trained, DL systems produce an output for any arbitrary numeric vector provided as input, regardless of whether it is within or outside the validity domain of the system under test. Hence, the quality of such systems is determined by the intersection between their validity domain and the regions where their outputs exhibit a misbehaviour.
@inproceedings{paperbot5836,
title = {Model-based exploration of the frontier of behaviours for deep learning system testing},
author = {Vincenzo Riccio and Paolo Tonella},
booktitle = {FSE},
year = {2020},
doi = {10.1145/3368089.3409730}
}