[Retracted] Research on Complex Classification Algorithm of Breast Cancer Chip Based on SVM-RFE Gene Feature Screening

Algorithm 1

SVM-RFE algorithm flow.

	Input: the training samples: , is the samples of d-dimensional space
	Category label:
	Initialization: feature sort , current feature index sequence
	Feature sorting: iterates in a loop until ;
	Step 1: obtain a new data sample according to the current feature:
	Step 2: train SVM with a new sample set to obtain support vector related parameters:
	Step 3: calculate the sorting factor
	Step 4: find out the feature with the smallest sorting criterion and add it to the feature sorting table:
	Step 5: remove the feature with the smallest sorting coefficient from the current remaining dataset:
	Output: sorted list of features.