1: \begin{abstract}

2: Support matrix machine (SMM) is a successful supervised classification model for matrix-type samples. Unlike support vector machines, it employs low-rank regularization on the regression matrix to effectively capture the intrinsic structure embedded in each input matrix.	

3: When solving a large-scale SMM, a major challenge arises from the potential increase in sample size, leading to substantial computational and storage burdens.

4: To address these issues, we design a semismooth Newton-CG (SNCG) based augmented Lagrangian method (ALM) for solving the SMM.

5: The ALM exhibits an asymptotic R-superlinear convergence if a strict complementarity condition is satisfied.

6: The SNCG method is employed to solve the ALM subproblems, achieving at least a superlinear convergence rate under the nonemptiness of an index set.

7: %under the condition of mild constraint nondegeneracy on the dual subproblems.

8: Furthermore, the sparsity of samples and the low-rank nature of solutions enable us to reduce the computational cost and storage demands for the Newton linear systems.

9: Additionally, we develop an adaptive sieving strategy that generates a solution path for the SMM by exploiting sample sparsity. The finite convergence of this strategy is also demonstrated.

10: Numerical experiments on both large-scale real and synthetic datasets validate the effectiveness of the proposed methods.

11: \keywords{Support matrix machine \and Sample sparsity \and Low rank \and Adaptive sieving \and Augmented Lagrangian method \and Semismooth Newton method}

12: % \PACS{PACS code1 \and PACS code2 \and more}

13: \subclass{90C06 \and 90C25 \and 90C90}

14: \end{abstract}