Research Article
Kernel Negative ε Dragging Linear Regression for Pattern Classification
Input: Training samples matrix ; Label matrix ; dragging coefficient matrix ; test | sample ; parameter ; | Output: the slack variable class label matrix ; predicted class for test sample ; | Initialization: ; | Calculate ; | Set threshold ; Set . | Repeat | Given , calculate . | Utilize , then calculate . | Until the absolute value of the difference between objective functions of two consecutive | loops is smaller than threshold . | For test sample , calculate . | If , then is classified into the th class. is the th entry of . | Output: the transformation matrix , . |
|