Breast cancer distinguishing method based on gaussian kernel function fuzzy noncorrelation distinguishing conversion

Abstract

The invention relates to a breast cancer distinguishing method based on gaussian kernel function fuzzy noncorrelation distinguishing conversion. A function theory is used for fuzzy noncorrelation distinguishing conversion, and quick and accurate distinguishing of breast cancer is realized. The breast cancer distinguishing method provided by the invention comprises the following steps of: firstly, conducting fuzzification on data; then calculating a first characteristic vector of an optimal distinguishing vector set; then calculating one group of optimal distinguishing vector set of the method provided by the invention, and achieving nonlinear conversion of fuzzy noncorrelation distinguishing conversion by utilizing the kernel function; and finally carrying out the nonlinear conversion on a breast cancer diagnosis data set so as to achieve correct distinguishing of the breast cancer. According to the breast cancer distinguishing method provided by the invention, solves the problem that linear impartibility is difficult to handle by the fuzzy noncorrelation distinguishing conversion, data of the breast cancer diagnosis data set is mapped to a high-dimensional characteristic space by utilizing nonlinear mapping, and a gassian kernel function implicit expression is used for achieving the calculation in the high-dimensional characteristic space, so that the problem of 'curse of dimensionality' can be avoided, nonlinear distinguishing information of the breast cancer diagnosis data set can be extracted, and classification accuracy rate is high.

Description

technical field [0001] The invention relates to the technical fields of pattern recognition and artificial intelligence, in particular to a breast cancer discrimination method based on Gaussian kernel function fuzzy non-correlation discrimination transformation. Background technique [0002] Breast cancer is the most common malignant tumor in women. According to statistics, an average of 1.3 million people around the world are newly diagnosed with breast cancer every year. Female breast cancer patients account for 30% of new malignant tumors in women, ranking first in the incidence of female malignant tumors. Breast cancer is also the most common malignant tumor among women in my country. Breast cancer can be detected by clinical history, physical examination, mammography, and contrast-enhanced ultrasonography. However, the definitive diagnosis of a breast mass must be made by fine-needle aspiration biopsy, central biopsy, or surgical excision. Among them, fine-needle asp...

AI Technical Summary

Problems solved by technology

However, the fuzzy non-correlation discriminant transformation method has difficulty in dealing with linear inseparable problems, and the processing effect is often unsatisfactory.

Images