An application of a neuro-fuzzy inference system for pattern classification of HDD defect distribution

Abstract

In the production of Hard Disk Drive (HDD), the pattern of HDD defect distribution gives the information about the defective component of HDD. This paper presents a pattern classification method for the HDD defect distribution. Five features of the pattern are developed and an Adaptive Neuro-Fuzzy Inference System (ANFIS) is applied to classify the input pattern. The ANFIS is trained with the sense of supervised learning. The trained ANFIS shows satisfactory results for classification of HDD defect distribution.