Windowed electroencephalographic signal classifier based on continuous neural networks with delays in the input
作者：
Highlights：
• An EEG signal classifier was developed based on a class of TDNN with delays appearing in the input signal.
• This characteristic was proposed to take into account the concept of signal windowing.
• The capability of a the TDNN to be employed as a EEG signal pattern classifier was tested.
• The training process was proposed based on the technique of Lyapunov–Krasovsky stability analysis.
• The correct classification of EEG signals attained a 99% of correct classification.
摘要
•An EEG signal classifier was developed based on a class of TDNN with delays appearing in the input signal.•This characteristic was proposed to take into account the concept of signal windowing.•The capability of a the TDNN to be employed as a EEG signal pattern classifier was tested.•The training process was proposed based on the technique of Lyapunov–Krasovsky stability analysis.•The correct classification of EEG signals attained a 99% of correct classification.
论文关键词：Time-delay neural networks,Electroencephalographic signals,Pattern classification,Lyapunov stability theory,Epileptic seizures
论文评审过程：Received 24 February 2016, Revised 24 June 2016, Accepted 3 August 2016, Available online 21 September 2016, Version of Record 6 October 2016.
论文官网地址：https://doi.org/10.1016/j.eswa.2016.08.020
