1: \begin{abstract}

2: In this paper, we develop a new alternating projection method to compute

3: nonnegative low rank matrix approximation for nonnegative matrices.

4: In the nonnegative low rank matrix approximation method,

5: the projection onto the manifold of fixed rank matrices can be expensive as the singular value

6: decomposition is required. We propose to use the

7: tangent space of the point in the manifold

8: to approximate the projection onto the manifold in order to reduce the computational cost.

9: We show that the sequence generated by the alternating projections onto the tangent spaces of the fixed rank matrices manifold and the nonnegative matrix manifold,

10: converge linearly to a point in the intersection of the two manifolds where the convergent point

11: is sufficiently close to optimal solutions. This convergence result based inexact projection onto the manifold

12: is new and is not studied in the literature.

13: Numerical examples in data clustering, pattern recognition and hyperspectral data

14: analysis are given to demonstrate that

15: the performance of the proposed method is better than that of nonnegative matrix factorization methods

16: in terms of computational time and accuracy.

17: \end{abstract}

18: