Self-organizing neural network topology preservation reinforcing method based on deep learning

Abstract

The invention provides a self-organizing neural network topology preservation reinforcing method based on deep learning. The method is used for settling a technical problem of requirement for improving self-organizing neural network topology preservation effect in prior art. The method comprises the steps of setting a network structure and a parameter of the self-organizing neural network, and normalizing input layer data; setting the number of input-layer neurons and the number of competition-layer neurons; performing rough adjustment on the weight vector of the competition-layer neurons of the self-organizing neural network, and obtaining a rough-adjusted competition-layer weight vector; by means of the rough-adjusted result, performing fine adjustment on the weight vector of the competition-layer neurons of the self-organizing neural network, and obtaining a fine-adjusted competition-layer weight vector; measuring the competition-layer weight vector, thereby obtaining a topology preservation reinforcing effect. The self-organizing neural network topology preservation reinforcing method has advantages of reducing difference between each competition-layer weight vector of the self-organizing neural network and an input sample, and improving topology preservation capability of the self-organizing neural network.

Description

technical field [0001] The invention belongs to the technical field of data processing, and relates to an enhancement method for self-organizing neural network topology preservation, in particular to an enhancement method for self-organizing neural network topology preservation based on deep learning in an unsupervised learning method, which can be used for high-dimensional data visualization, compression, and mining. Background technique [0002] In the process of processing high-dimensional data, data visualization and dimensionality reduction are of great significance. Visualization and dimensionality reduction of high-dimensional data can help us observe the approximate distribution of the data, and then further process the data. In this process, in order to ensure that the dimensionality reduction and visualization data can accurately reflect the distribution of the original data, the processing process should be topologically preserved, that is, the topology structure...

AI Technical Summary

Problems solved by technology

[0008] The technical idea of ​​the present invention is: through the method of deep learning, introduce more competitive layers, so that the network can learn the topology structure of the original data more fully; Ability, two processes of rough tuning and fine tuning are designed, without introducing too many hyperparameters, it is used to solve the problem of poor topology preservation ability of the self-organizing neural network in the prior art when processing small sample data

Images