COMPARISON OF NEURAL NETWORK METHODS OF DATA PREPROCESSING IN SOLVING PROBLEMS OF ANALYSIS, DIAGNOSIS, AND CLASSIFICATION OF DEFECTS OF RADIO-ELECTRONIC EQUIPMENT
B.P. Ivanenko, S.A. Klestov, V.I. Syryamkin
National Research Tomsk State University, Tomsk, Russia
Keywords: flaw detection, neural networks, splines, regularization
Abstract
We analyses the issues with processing data obtained from a 3d X-ray microtomograph when dealing with problem diagnosis and classification of defects in radio-electronic devices. It is proposed to use neural network methods and regularizing splines for solving the problem. A comparative analysis of neural network and spline methods is carried out in solving problems of recovering heavily noisy signals. The effectiveness of the proposed approach is studied by numerical simulation methods and in the processing of real data.
|
