P‐53: Analysis of dead pixel origins and potential defect prediction using machine learning with tabular data

Producing defect‐free products is one of the essential tasks of a manufacturing company. In this paper, the causes of dead pixels are analyzed and the coordinates of dead pixels are predicted using machine learning to solve problems of dead pixels. Unlike the existing method of analyzing defects using images, tabular data where each element contains significance were used, and detailed description of suggested models are depicted. Experiments ensured the consistency of the proposed method.