Optimization of unknown defect rejection for automatic defect classification

A non-transitory computer readable storage medium having instructions that, when executed by a processing device, cause the processing device to perform operations comprising:

• storing a definition of a region in a feature space, wherein the definition is associated with a class of defects and comprises a kernel function comprising a parameter, wherein the parameter determines a shape of the region;
• receiving a confidence threshold for automatic classification of at least one defect associated with the class;
• selecting a value of the parameter associated with the confidence threshold;
• receiving inspection data for a plurality of defects detected in one or more samples under inspection;
• and
• automatically classifying the plurality of defects for the class using the kernel function and the selected value of the parameter.