Few-sample learning classifier construction method based on unbalanced data

Abstract

The invention relates to a few-sample learning classifier construction method based on unbalanced data, and belongs to the technical field of computer data classification. The method comprises the following steps: firstly, designing a twinning parallel full connection network for feature learning of input sample pairs according to primary learning and few sample learning characteristics of a twinning neural network; and then, processing the imbalance problem of the input sample pairs by using a cost sensitive optimizer, designing an expected error classification cost function according to different error classification costs, and integrating the expected error classification cost function into a network parameter optimization algorithm for adjusting class imbalance classification weights.According to the method, a better classification result can be obtained under unbalanced, high-dimensional and limited target data sets, and the classification performance is more stable.

Description

technical field [0001] The invention belongs to the technical field of computer data classification, and relates to a method for constructing a few-sample learning classifier based on unbalanced data. Background technique [0002] Data classification is one of the key research contents of data mining. It uses valuable available data to classify unknown data, and aims to explore the hidden relationship between variables and classes. Most of the current data-driven machine learning classification algorithms assume that the proportion of samples of each type of target data is the same, but in the actual binary classification task, the target data is often unbalanced, high-dimensional and limited, in this case , conventional machine learning classification algorithms are difficult to obtain good classification results. [0003] For class unbalanced data, Piri et al. in the article "S.Piri, D.Delen, T.Liu, A synthetic informative minority over-sampling (simo) algorithm leveragin...

AI Technical Summary

Problems solved by technology

[0007] In view of this, the object of the present invention is to provide a method for constructing a few-sample learning classifier based on unbalanced data, which is used to solve the problem that unbalanced, high-dimensional and limited target data cannot be uniformly processed for binary classification, so that Achieve better classification results with unbalanced, high-dimensional and limited target datasets

Images