Research Article
A Novel Convex Clustering Method for High-Dimensional Data Using Semiproximal ADMM
(1) | Initialization Let and . Given a start point , for | (2) | repeat | (3) | Step 1. Update according to (25) | (4) | Step 2. Update according to (28), (29) and (30) | (5) | Step 3. Update the Lagrangian multipliers by (21) | (6) | until Stopping criterion is satisfied. |
|