首页|New Machine Learning Study Findings Reported from East Carolina University (Impr oving Early Fault Detection in Machine Learning Systems Using Data Diversity-Dri ven Metamorphic Relation Prioritization)
New Machine Learning Study Findings Reported from East Carolina University (Impr oving Early Fault Detection in Machine Learning Systems Using Data Diversity-Dri ven Metamorphic Relation Prioritization)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-New study results on artificial intell igence have been published. According to news reporting originating from Greenvi lle, North Carolina, by NewsRx correspondents, research stated, "Metamorphic tes ting is a valuable approach to verifying machine learning programs where traditi onal oracles are unavailable or difficult to apply." Our news journalists obtained a quote from the research from East Carolina Unive rsity: "This paper proposes atechnique to prioritize metamorphic relations (MRs ) in metamorphic testing for machine learning and deep learning systems, aiming to enhance early fault detection. We introduce five metrics based on diversity i n source and follow-up test cases to prioritize MRs. The effectiveness of our pr oposed prioritization methods is evaluated on three machine learning and one dee p learning algorithm implementation. We compare our approach against random-base d, fault-based, and neuron activation coverage-based MR ordering. The results sh ow that our data diversity-based prioritization performs comparably to fault-bas ed prioritization, reducing fault detection time by up to 62% comp ared to random MR execution. Our proposed metrics outperformed neuron activation coverage-based prioritization, providing 5-550% higher fault dete ction effectiveness."
East Carolina UniversityGreenvilleNo rth CarolinaUnited StatesNorth and Central AmericaCyborgsEmerging Techno logiesMachine Learning
NETL
NSTL
