Artificial Neural Networks in the Problem of Classification of Defects of Electronic Devices

摘要

The problem of identification and classification of mechanical defects of electronic devices has been studied, and a solution of this problem using an artificial neural network has been proposed. It is shown how artificial neural networks may be used to solve the problem faster and ensure at the same time the required validity. Based on experiments several neural networks have been developed under various input conditions that were trained and put in operation.