Multi-instance multi-label scene classification method based on multinuclear fusion

Abstract

The invention relates to a multi-instance multi-label scene classification method based on multinuclear fusion. The method comprises the following steps of inputting one multi-instance multi-label data set and splitting into a multi-instance data set and a multi-label data set; using different thresholds to establish a correlation matrix for each package in the multi-instance data set; according to the obtained correlation matrix, calculating a basic nuclear function between each two multi-instance data packages under a same threshold, wherein a basic nuclear function value forms a basic nuclear matrix; carrying out convex combination on element values of a same position in the basic nuclear matrixes under the different thresholds so as to obtain a multinuclear matrix; using the multi-label data set to carry out training so as to obtain a plurality of multinuclear SVM classifiers. The plurality of multinuclear SVM classifiers are used for predicting a label set of an unknown multi-instance data package so as to realize scene classification. By using the multi-instance multi-label scene classification method based on the multinuclear fusion, scene classification accuracy is increased. The invention also relates to a multi-instance multi-label scene classification system based on the multinuclear fusion.

Description

technical field [0001] The invention relates to the technical field of machine learning, in particular to a multi-instance multi-label scene classification method based on multi-core fusion. Background technique [0002] Multi-instance learning is a learning method evolved from supervised learning. It was first proposed in the 1990s when people were studying drug activity. It regards each pharmaceutical molecule as a package, and each isomer of the molecule is viewed as As an example in the package, if the molecule has an isomer that is suitable for pharmaceuticals, then the package corresponding to the molecule is marked as a positive package, otherwise it is marked as a negative package. Through this method, a learning system is finally constructed, and then the Molecules that are known to be suitable or unsuitable for pharmaceuticals are learned to correctly predict whether other new molecules are suitable for pharmaceuticals. Since then, multi-instance learning has been...

AI Technical Summary

Problems solved by technology

[0004] Although the above algorithms have achieved good results in solving the multi-instance multi-label problem, they do not consider the correlation of examples in the bag.

In many practical applications, especially in scene classification problems, the independence assumption of examples is difficult to guarantee, which will lead to unsatisfactory classification results.

Images