Neural network construction, training and recognition method, system, and storage medium

Abstract

The invention discloses a neural network construction, training and recognition method, a system, and a storage medium. Due to discontinuity of a neuron internal state variable and an error function,an input layer pulse sequence is constructed by using a convolution kernel function of a time pulse. In the hidden layer, the membrane voltage of synaptic neurons is solved by utilizing an electric leakage-integration-excitation neuron model. When the membrane voltage exceeds a predefined threshold value, one pulse is excited to be transmitted to the next layer. In a network output layer, trainingerrors are calculated through a difference value between a desired pulse sequence and a real pulse sequence. After that, a Widrow-Hoff (WH) learning rule is simulated to carry out gradient error reduction and back propagation. The neural network model has strong capability of processing complex time series data (such as target recognition and voice recognition in a video), and is helpful for realizing the application of brain-like calculation in practice.

Description

technical field [0001] The invention belongs to the technical field of machine learning, pattern classification and recognition, and in particular relates to a multi-layer multi-pulse neural network model construction method and training method for dealing with complex spatio-temporal pattern classification problems, as well as a pattern recognition system and a sample recognition method based on the model. and storage media. Background technique [0002] Since the plasticity between neurons makes the biological nervous system have a strong ability to learn and adapt to the environment, it is extremely important to consider the adjustment ability of synapses caused by changes in the external environment and neural processes in modeling. Synaptic Weights (Synaptic Weights) define the strength of the connection between two neurons. Hebb proposed the first hypothesis about the modification of synaptic weights. The learning algorithm based on this hypothesis can be summarized as...

AI Technical Summary

Problems solved by technology

[0005] According to the above, in order to solve the problems existing in the classification of multi-layer multi-pulse neural network learning algorithms in processing complex spatio-temporal patterns, the complex spatio-temporal patterns are patterns that contain both time information and spatial information, the present invention proposes a neural network Network construction, training, identification method and system, storage medium

Images