Nuisance Rate Improvement of E-beam Defect Classification

The proposed paper presents a case study describing how e-beam defect classification nuisance rate (NR) can be improved by the implementation of a new machine learning classification process in HMI e-Manager even for difficult data (feature boundary is overlay). This is important because low nuisance rate is an importance metric to measure the e-beam defect classification performance and it is usually difficult to obtain the low nuisance rate, especially for difficult defect dataset. Our machine learning (not a deep learning) multiple-phase classification results show that it is an effective way to improve the E-beam defect classification nuisance rate.