Abstract
A novel multi-layer multi-spiking neural network (MMSNN) model sends information from one neuron to the next through multiple synapses in different spikes. When a suitable training method is unavailable, there are numerous complications in training spike neural networks. The convergence rate is slow when gradient descent algorithms are trained on multi-spike neural networks. Meanwhile, the spiking network is likely to perform poorly in classifying electroencephalographic (EEG) datasets due to its high propensity to fall within the local minima. With the help of a 16-channel electrode, the Concealed Information Test created the EEG dataset. The Mini Batch Stochastic Gradient Descent (MBSGD) built the batch size and cut the number of iterations to speed up the convergence rate. It can, however, avoid the problem of moving into the local minimum more effectively. This paper describes an MBSGD algorithm for MMSNN model training that addresses local minima and convergence rate issues. This paper presents an MBSGD-MMSNN model for classifying EEG data. The MBSGD algorithm adjusts the weight parameters of the MMSNN model based on the batch size until the error is less than the required value. As a result, we achieve a faster convergence rate and avoid issues with local minima. The performance of the proposed MBSGD-MMSNN model is compared to that of existing meta-heuristic algorithms. The proposed MBSGD-MMSNN model outperformed other state-of-the-art algorithms regarding the accuracy and learning performance.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Data availability
Enquiries about data availability should be directed to the authors.
References
Akhavan A, Moradi MH, Vand SR (2017) Subject-based discriminative sparse representation model for detection of concealed information. Comput Methods Programs Biomed 143:25–33
Arasteh A, Moradi MH, Janghorbani A (2016) A novel method based on empirical mode decomposition for P300-based detection of deception. IEEE Trans Inf For Secur 11(11):2584–2593
Bablani A, Edla DR, Tripathi D, Cheruku R (2019) Survey on brain-computer interface: an emerging computational intelligence paradigm. ACM Comput Surv 52(1):1–32
Bablani A, Edla DR, Tripathi D, Dodia S, Chintala S (2019) A synergistic concealed information test with a novel approach for EEG channel selection and SVM parameter optimization. IEEE Trans Inf Forensics Secur 14(11):3057–3068
Bablani A, Edla DR, Tripathi D (2019) An efficient concealed information test: EEG feature extraction and ensemble classification for lie identification. Mach Vis Appl 30(5):813–832
Booij O, Tat Nguyen H (2005) A gradient descent rule for spiking neurons emitting multiple spikes. Inf Process Lett 95(6):552–558
Dodia S, Edla DR, Bablani A, Ramesh D, Kuppili V (2019) An efficient EEG based deceit identification test using wavelet packet transform and linear discriminant analysis. J Neurosci Methods 314:31–40
Dodia S, Edla DR, Bablani A, Cheruku R (2020) Lie detection using extreme learning machine: a concealed information test based on short-time Fourier transform and binary bat optimization using a novel fitness function. Comput Intell 36(2):637–658
Farahani ED, Moradi MH (2017) Multimodal detection of concealed information using genetic-SVM classifier with strict validation structure. Inform Med Unlocked 9:58–67
Farwell LA, Donchin E (1991) The truth will out: interrogative polygraphy (‘lie detection’) with event-related brain potentials. Psychophysiology 28(5):531–547
Gao J, Lu L, Yang Y, Yu G, Na L, Rao N (2012) A novel concealed information test method based on independent component analysis and support vector machine. Clin EEG Neurosci 43:54–63
Ghosh-Dastidar S, Adeli H (2009) A new supervised learning algorithm for multiple spiking neural networks with application in epilepsy and seizure detection. Neural Netw 22(10):1419–1431
Hodgkin AL, Huxley AF (1952) A quantitative description of membrane current and its application to conduction and excitation in nerve. J Physiol 117:500–544
Hu T, Lin X, Wang X, Du P (2022) Supervised learning algorithm based on spike optimization mechanism for multilayer spiking neural networks. Int J Mach Learn Cybern. https://doi.org/10.1007/s13042-021-01500-8
Izhikevich EM (2003) Simple model of spiking neurons. IEEE Trans Neural Netw 14(6):1569–1572
Jenke R, Peer A, Buss M (2014) Feature extraction and selection for emotion recognition from EEG. IEEE Trans Affective Comput 5(3):327–339
Jung TP et al (2000) Removing electroencephalographic artifacts by blind source separation. Psychophysiology 37(2):163–178
Li Y, Guo Y, Zhang S, Deng S, Hai Y, Gu S (2021) Differentiable spike: rethinking gradient-descent for training spiking neural networks. Adv Neural Inf Process Syst 34:23426
Luo Y, Shen H, Cao X, Wang T, Feng Q, Tan Z (2022) Conversion of Siamese networks to spiking neural networks for energy-efficient object tracking. Neural Comput Appl 34(12):9967–9982
Meixner JB, Rosenfeld JP (2011) A mock terrorism application of the P300-based concealed information test. Psychophysiology 48(2):149–154
Ramadan RA, Vasilakos AV (2017) Brain computer interface: control signals review. Neurocomputing 223:26–44
Rosenfeld JP, Labkovsky E, Winograd M, Lui MA, Vandenboom C, Chedid E (2008) The complex trial protocol (CTP): a new, countermeasure-resistant, accurate, P300-based method for detection of concealed information. Psychophysiology 45(6):906–919
Sung M, Kim Y (2020) Training spiking neural networks with an adaptive leaky integrate-and-fire neuron. In: 2020 IEEE international conference on consumer electronics-Asia (ICCE-Asia), pp. 1–2, IEEE
Wang D, Miao D, Blohm G (2013) A new method for EEG-based concealed information test. IEEE Trans Inf For Secur 8(3):520–527
Funding
The authors have not disclosed any funding.
Author information
Authors and Affiliations
Contributions
All the authors contributed equally to this research.
Corresponding author
Ethics declarations
Conflict of interest
No Conflict of interest.
Ethical approval
This article contains no studies with human or animal participants conducted by any of the authors.
Human or animal rights
No involving human participants and/or animals.
Informed consent
Not applicable.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Ramesh, M., Revoori, S., Edla, D.R. et al. A novel multi-layer multi-spiking neural network for EEG signal classification using Mini Batch SGD. Soft Comput 27, 9877–9890 (2023). https://doi.org/10.1007/s00500-023-08404-5
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00500-023-08404-5