Linear Discriminative Learning for Image Classification
Algorithm 1
Proposed method.
Input: A-unit -normalized data matrix; unit -normalized query samples matrix and , regularity parameter
Output: return labels of the query samples
for to n do
Use equation (19) and determine the similarity matrix Q; estimate weighted matrix
Compute coefficients using equation (20), to estimate the residual and update as
Update and solve the objective function in equation (11) to obtain linear representation ; compute coding error against each class using equation (21); assign the class label to query sample according to equation (22)
