Prediction of the RFID Identification Rate Based on the Neighborhood Rough Set and Random Forest for Robot Application Scenarios

Algorithm 1

Forward greedy attribute reduction.

	Input: Neighborhood decision system W = (U, C ∪ D, V, f, ε), ε is neighborhood threshold, T is temporary subset.
	Output: Reduction subset R.
	Steps:
(1)
(2)	, Calculate the attribute importance sig of the sample ;
(3)	, Find the attribute with the most important attribute of attribute reduction attribute subset ;
(4)	If ()
(5)	, ;
(6)	Else
(7)	return R;
(8)	End for