1: \begin{abstract}

2: In this paper we analyze several new methods for solving nonconvex

3: optimization problems with the objective function consisting of a sum of

4: two  terms: one is nonconvex and smooth, and another is convex but

5: simple and its structure is known. Further, we consider both cases:

6: unconstrained and linearly constrained nonconvex problems. For

7: optimization problems of the above structure,  we propose random

8: coordinate descent algorithms and analyze  their convergence

9: properties. For the general case, when the objective function is

10: nonconvex and composite we prove asymptotic convergence for the

11: sequences generated by our algorithms to stationary points  and

12: sublinear rate of convergence in expectation for some optimality

13: measure. We also present

14: extensive numerical experiments for evaluating the performance of

15: our algorithms in comparison with state-of-the-art methods.

16: \end{abstract}

17: