一句话：提出 TERA，优化机器学习项目中随机回归测试的超参数配置，降低回归测试成本。
机器学习算法具有随机性，且其精度与运行时间受超参数控制，这使得 ML 工具与库的测试颇具挑战。回归测试中，开发者通常手动选择超参数，往往过于保守，导致测试执行缓慢并推高回归测试成本。论文提出 TERA，面向机器学习项目优化随机回归测试，试图缓解手工、保守超参数配置带来的开销。摘要仅给出问题动机，未说明 TERA 的具体技术手段与实验评估；其核心目标是在不显著损害测试有效性的前提下降低随机回归测试的执行成本。
The stochastic nature of many Machine Learning (ML) algorithms makes testing of ML tools and libraries challenging. ML algorithms allow a developer to control their accuracy and run-time through a set of hyper-parameters, which are typically manually selected in tests. This choice is often too conservative and leads to slow test executions, thereby increasing the cost of regression testing.
DOI 原文 ·
@inproceedings{paperbot6664,
title = {TERA: optimizing stochastic regression tests in machine learning projects},
author = {Saikat Dutta and Jeeva Selvam and Aryaman Jain and Sasa Misailovic},
booktitle = {ISSTA},
year = {2021},
doi = {10.1145/3460319.3464844}
}