Label Distribution Learning by Regularized Sample Self-Representation

Algorithm 2

Regularized sample self-representation by -norm (RSSR-LDL21).

Input: Train matrix , test data , the number of iterations and the regularization parameter .

Output: The corresponding predicted label distribution of test data , .

Initialize which is an identity matrix, set ; // Initialization

for   to Iter do.

, // calculate

, // calculate

end for

  ; // Removing the matrix part of matrix .

  ; // Using the transition matrix to represent the distribution of the predicted labels.

  ; // Because of and , the normalized , .