Feature selection approach based on neural-fuzzy network

Abstract

The invention provides a feature selection approach based on a neural-fuzzy network. The feature selection approach includes the following steps of: the first step, adjusting parameters of fuzzy membership functions to obtain a membership function set shown in the description, based on an x training neural fuzzy network, wherein [mu]fimj represents an mj-th membership function of features fi; the second step, calculating the output oq of the neural-fuzzy network when the input is xq; the third step, modifying the neural-fuzzy network to allow a fuzzy mapping layer to map the xq as that shown in the description, wherein the values of all the membership functions of the fi are identically equal to 0.5, and at this moment, the input of the neural-fuzzy network is that shown in the description; the fourth step, calculating measurement values FQJi shown in the description of the feature fi; and the fifth step, sorting the feature measurement values FQJi in a descending order.

Description

technical field [0001] The present invention relates to the technical field of pattern recognition, more specifically, the present invention relates to a feature selection method based on neural fuzzy network. Background technique [0002] Feature selection is a key problem in techniques such as pattern recognition. In recent years, the use of artificial neural networks for feature selection is a hot issue. Through artificial neural networks with good learning performance, the importance of individual features or feature subsets can be inferred, but there are some problems. [0003] In the field of feature selection, most feature selection based on artificial neural networks can be regarded as a special case of network pruning algorithm, the difference is that the input nodes are pruned instead of hidden layer nodes or weights. At present, the widely used method is to use the change amount between the network output before and after pruning as the measure of the importance...

AI Technical Summary

Problems solved by technology

However, since the definition of its membership function is actually done before network learning, the membership mapping may still cause the mapped data to be distorted

Images