Fast Prediction with Sparse Multikernel LS-SVR Using Multiple Relevant Time Series and Its Application in Avionics System

Pseudocode 1

Pseudocode of the proposed pruning method.

parameters: , , support vectors set

Initialize: , equals to the LS-SVR parameter

(1) choose the initial BV: and ;

(2) establish the initial BVS: ;

(3) get new support vector ;

(4) approximate the linear dependence compute via (23)

     if 

       ,

       ,

       add into the BVS,

     else

       drop ,

(5) go back to (3).