Data-driven associated evidence fusion method based on PCA (Principal Component Analysis)

The invention relates to a data-driven association evidence fusion method based on PCA, and the method comprises the following steps: obtaining classified and recognized original data, and obtaining an original data set; based on a PCA method, obtaining independent principal components corresponding to the original data so as to obtain a new data set containing a plurality of principal components; according to the new data set, training to obtain a data driving model corresponding to each principal component; constructing a basic probability assignment function corresponding to each principal component according to the principal component value determined by the attribute value of the sample to be identified; and fusing the basic probability assignment functions of the principal components, and judging to obtain a final recognition result. Compared with the prior art, principal component analysis (PCA) is utilized to highlight the main features of the data set, some associated redundant information is rejecte