1: \begin{abstract}

2:  This paper considers optimization problems on the Stiefel manifold $X^{\msT}X=I_p$, where $X\in \mathbb{R}^{n \times p}$ is the

3:  variable and $I_p$ is the $p$-by-$p$ identity matrix. A framework of constraint preserving update schemes is proposed by decomposing each feasible point into the range space  of $X$ and the null space of $X^{\msT}$. While this general framework can unify many existing schemes, a new update scheme with low complexity cost is also discovered.  Then we study a feasible Barzilai-Borwein-like method under the new update scheme. The global convergence of  the method is established with an adaptive nonmonotone line search. The numerical tests on the  nearest low-rank correlation matrix problem, the Kohn-Sham total energy minimization  and a specific problem from statistics demonstrate the efficiency of the new method. In particular, the new method performs \rev{remarkably well for the nearest low-rank correlation matrix problem in terms of  \rrev{speed} and solution quality} and is considerably competitive with the widely used SCF iteration for the Kohn-Sham total energy minimization.

4:  %and the returned solution {\color{red}significantly  better} than some state-of-the-art algorithms for the nearest low-rank correlation matrix problem and

5: 

6: \end{abstract}