Title :
Method of spectrogram recognition by artificial neural network
Author :
Orlikov, N.L. ; Petrakov, R.S.
Author_Institution :
Tomsk State Univ. of Control Syst. & Radioelectronics, Russia
Abstract :
Methods of spectrogram recognition by artificial neural network based on sigmoid, linear and stepped functions are considered. It is shown that the application of neural technologies essentially simplifies the interfacing of technological devices with computers. In the spectrum analysis problem, using the artificial neural network in the case of spectrogram imposition gives more exact results than statistical methods
Keywords :
neural nets; pattern recognition; spectral analysis; spectroscopy; artificial neural network; linear functions; neural technologies; sigmoid functions; spectrogram recognition; spectrum analysis; statistical methods; stepped functions; Artificial neural networks; Control systems; Decoding; Electronic mail; Gases; Hardware; Neural networks; Neurons; Radio control; Spectrogram;
Conference_Titel :
Modern Communication Technologies, 2001. SIBCOM-2001. The IEEE-Siberian Workshop of Students and Young Researchers
Conference_Location :
Tomsk
Print_ISBN :
0-7803-6611-5
DOI :
10.1109/SIBCOM.2001.977514
