Next Article in Journal
Internet of Measurement Things Architecture: Proof of Concept with Scope of Accreditation
Previous Article in Journal
Classification of Dried Strawberry by the Analysis of the Acoustic Sound with Artificial Neural Networks
Open AccessArticle

Competitive Learning in a Spiking Neural Network: Towards an Intelligent Pattern Classifier

Neurotechnology Department, Lobachevsky State University of Nizhny Novgorod, 603950 Nizhny Novgorod, Russia
*
Author to whom correspondence should be addressed.
Sensors 2020, 20(2), 500; https://doi.org/10.3390/s20020500
Received: 3 December 2019 / Revised: 10 January 2020 / Accepted: 14 January 2020 / Published: 16 January 2020
(This article belongs to the Section Biomedical Sensors)
One of the modern trends in the design of human–machine interfaces (HMI) is to involve the so called spiking neuron networks (SNNs) in signal processing. The SNNs can be trained by simple and efficient biologically inspired algorithms. In particular, we have shown that sensory neurons in the input layer of SNNs can simultaneously encode the input signal based both on the spiking frequency rate and on varying the latency in generating spikes. In the case of such mixed temporal-rate coding, the SNN should implement learning working properly for both types of coding. Based on this, we investigate how a single neuron can be trained with pure rate and temporal patterns, and then build a universal SNN that is trained using mixed coding. In particular, we study Hebbian and competitive learning in SNN in the context of temporal and rate coding problems. We show that the use of Hebbian learning through pair-based and triplet-based spike timing-dependent plasticity (STDP) rule is accomplishable for temporal coding, but not for rate coding. Synaptic competition inducing depression of poorly used synapses is required to ensure a neural selectivity in the rate coding. This kind of competition can be implemented by the so-called forgetting function that is dependent on neuron activity. We show that coherent use of the triplet-based STDP and synaptic competition with the forgetting function is sufficient for the rate coding. Next, we propose a SNN capable of classifying electromyographical (EMG) patterns using an unsupervised learning procedure. The neuron competition achieved via lateral inhibition ensures the “winner takes all” principle among classifier neurons. The SNN also provides gradual output response dependent on muscular contraction strength. Furthermore, we modify the SNN to implement a supervised learning method based on stimulation of the target classifier neuron synchronously with the network input. In a problem of discrimination of three EMG patterns, the SNN with supervised learning shows median accuracy 99.5% that is close to the result demonstrated by multi-layer perceptron learned by back propagation of an error algorithm. View Full-Text
Keywords: EMG interface; STDP; pair-based STDP; triplet-based STDP; temporal coding; rate coding; synaptic competition; neural competition; lateral inhibition EMG interface; STDP; pair-based STDP; triplet-based STDP; temporal coding; rate coding; synaptic competition; neural competition; lateral inhibition
Show Figures

Figure 1

MDPI and ACS Style

Lobov, S.A.; Chernyshov, A.V.; Krilova, N.P.; Shamshin, M.O.; Kazantsev, V.B. Competitive Learning in a Spiking Neural Network: Towards an Intelligent Pattern Classifier. Sensors 2020, 20, 500.

Show more citation formats Show less citations formats
Note that from the first issue of 2016, MDPI journals use article numbers instead of page numbers. See further details here.

Article Access Map by Country/Region

1
  • Supplementary File 1:

    ZIP-Document (ZIP, 23570 KB)

  • Externally hosted supplementary file 1
    Link: https://drive.google.com/file/d/14I5--V25MFQHpoqU9kJjkOvVefymycZb
    Description: Video S1: Unsupervised SNN learning The output neurons in the process of learning become selective to different EMG patterns generated by the muscles during a) wrist extension, b) wrist flexion, c) rest. It is impossible to predict which neuron will be responsible for which gesture. At the end of learning, we show that trained neuron has different couplings depending on what signals it responds on. The degree of grayscale of coupling is proportional to the value of weight.
  • Externally hosted supplementary file 2
    Link: https://drive.google.com/file/d/1tOHdpzYv8Ndw9UpxO1md6NViy3HFqK2A
    Description: Video S2: Supervised SNN learning Supervised learning is stimulation of the target neuron simultaneously with the generation of the corresponding EMG pattern. We would like to achieve the following correspondences of output neurons: a) the left neuron – the movement of the palm to the left, i.e. wrist flexion, b) the middle neuron – rest, c) the right neuron – the movement of the palm to the right, i.e. wrist extension.
Back to TopTop