An Efficient Tensor Completion Method Combining Matrix Factorization and Smoothness

Algorithm 2

The Proposed ADMM-based Algorithm for Solving the SMF Model in (11).

Input: The initialized tensor , index set , parameters .

Output: The completed tensor .

1: initialize:

2: while not converged and do

3:  for to 3 do

4:    via (17)

5:    via (22)

6:  end for

7:   via (24)

8:   via (27)

9:   via (29)

10:   via (30)

11: end while

12: return