An Efficient and Fast Model Reduced Kernel KNN for Human Activity Recognition

Algorithm 3

The working process of RK-KNN.

	Require:
	Training data matrix ;
	Testing data matrix ;
	Number of training data ;
	Parameter of RK-KNN ;
	Kernel parameter ;
	The percentage of selected data for reducing kernel matrix ;
	Ensure:
Prediction class .
	Reduced Kernel Computation Part:
(1)	Select P percentage samples for each class from the training data as ;
(2)	Calculate the reduced kernel matrix for training features by equation (9);
(3)	Calculate the reduced kernel matrix for testing features by equation (10);
	Loop:
(4)	fordo
(5)	fordo
(6)	Calculate the distance between and by equation (1);
(7)	end for
(8)	Sort the distance in the ascending order;
(9)	Pick the top K vectors from the sorted collection as an index;
(10)	Set the forecasting the class label based on the most frequent class of processed index.
(11)	end for
	return