Extension neural network pattern recognition method based on priori knowledge

Abstract

The invention discloses an extension neural network pattern recognition method based on priori knowledge. The method includes the following steps that (1) a training sample set and a knowledge base are prepared; (2) an initial weight value of an extension neural network is determined according to training samples and the priori knowledge; (3) the extension neural network can be trained by the utilization of the training samples, if a training process is converged or the total error rate reaches a preset value, training is stopped, and a weight value vector, after the training, of the extension neural network is kept, and otherwise the training is continued; (4) the trained extension neural network is used for performing pattern recognition until recognition of all objects to be recognized is completed. According to the extension neural network pattern recognition method, under the common driving of the priori knowledge and the training samples, learning of the extension neural network is guided, training and learning of the extension neural network are completed, the learning burden of the extension neural network is relieved, the performance of the extension neural network is effectively improved, training time is shortened, and recognition accuracy is improved.

Description

technical field [0001] The invention relates to the field of neural network pattern recognition, in particular to an extension neural network pattern recognition method driven by prior knowledge and training samples. Background technique [0002] Extension Neural Network (ENN) is the product of the organic combination of extension theory and neural network technology. Extension neural network is another new type of network after fuzzy neural network, genetic neural network and evolutionary neural network. Its appearance and development can not only expand the further application of extension theory itself, but also will promote the Further development of technology and intelligent computing. Currently, M.H.Wang's extension neural network "Extension neural network and its applications" (Neural Networks, 2003, Volume 16, Issue 5) proposed by M.H.Wang is the most widely used extension neural network model. It has a remarkable effect on solving problems such as classification,...

AI Technical Summary

Problems solved by technology

However, obtaining high-quality training samples is not an easy task, because although the observed data (sample data) and unseen data come from the same distribution, the number of observed data is usually limited and cannot describe the distribution of the original data well; In addition, although the number of sample data is sometimes sufficient, there are a large amount of data with little information, or a large amount of redundant data with similar information; at the same time, in the process of sample data generation, especially training samples obtained in harsh environments, It is likely that there will be various noises

Therefore, in the case of poor quality training samples, the performance of the extension neural network drops sharply. Therefore, we need to solve the practical problem of how to improve the performance of the extension neural network in the environment of poor samples.

[0004] In the actual application process, on the one hand, it is sometimes difficult to obtain high-quality sample data, but on the other hand, although some systems are so complex that it is impossible to understand all their internal structures, it is usually possible to have a certain understanding of certain process mechanisms, know Some prior knowledge information of these processes, therefore, it is hoped to make full use of this information, establish the connection between observed data and unseen data, and combine the method of empirical modeling to establish the model

Images