Kernel-Based Aggregating Learning System for Online Portfolio Optimization

Algorithm 2

Solving weights of aggregating estimator.

Input: Given the parameter e, learning rate η, and initial vector .
(1)	for s = 1 to t do
(2)	Calculate the final price prediction ;
(3)	Receive and incur loss ;
(4)	Let the gradient , update the weight vector .
(5)	end for
Output: Final price relative prediction .