Feature selection-based food dopant weight visual analysis method

A food dopant visual analysis method based on feature selection comprises the following steps: firstly, cleaning and preprocessing food inspection data, constructing a sample-dopant data set, then taking the processed data as the input of a model in feature selection, calculating the weight value of a dopant, and calculating the weight value of the dopant; then storing information such as a sample classification result, an evaluation index and a model structure in a learning process; and calculating a correlation measure between the features. Further designing a data visualization view to display the data; and finally, linkage interaction among multiple views is carried out to support a user to obtain an optimal feature combination according to information iterative analysis. The method is simple in operation and friendly in interface, and a user can obtain insights of weights of unqualified samples and dopants thereof in food inspection through the system without more domain knowledge. 基于特征选择的食品掺杂物可视分析方法，首先对