RnRTD: Intelligent Approach Based on the Relationship-Driven Neural Network and Restricted Tensor Decomposition for Multiple Accusation Judgment in Legal Cases

Algorithm 3

The framework of RnRTD algorithm.

	Input: , where represents the legal cases and represents the category of legal case corresponding to according to judgment results. The size of η, , , , , , , , , and , where , , , and
	Output: The optimal restricted tensor η, parameters of rdRNN , , , , and , , , , .
	Initialize the restricted tensor η, parameters of rdRNN , , , , and , , , ,
	for to do
	Set to zero;
	while do
	Set , , , , , and , , , , to zero;
	for to do
	;
	;
	, , , , , , , , , , ;
	end
η, , , , , , , , , , ;
	end
	end
	return η, , , , , and , , , , ;