1: \begin{abstract} 

2:   Leading eigenvalue problems for large scale matrices arise in many

3:   applications. Coordinate-wise descent methods are considered in

4:   this work for such problems based on a reformulation of the leading

5:   eigenvalue problem as a non-convex optimization

6:   problem. The convergence of several coordinate-wise methods is analyzed

7:   and compared. Numerical examples of applications to quantum many-body

8:   problems demonstrate the efficiency and provide benchmarks of the

9:   proposed coordinate-wise descent methods.

10: \end{abstract}

11: