1: \begin{abstract}

2:        This paper establishes the iteration-complexity of proximal bundle methods for solving

3:  hybrid (i.e., a blend of smooth and nonsmooth) weakly convex composite optimization (HWC-CO) problems.

4:  This is done in a unified manner by considering a proximal bundle framework (PBF) based on a generic bundle update scheme which includes various well-known bundle update schemes.

5:  In contrast to

6:  other well-known stationary conditions 

7:  % (e.g., the Moreau  stationary one)

8: in the context of

9: HWC-CO, PBF uses a new stationarity measure

10: % in the context of HWC-CO

11: which is easily verifiable and,

12: at the same time, implies any of the former ones.

13:  

14:  

15:        

16:        % a proximal bundle framework (PBF) for hybrid weakly convex composite optimization (HWC-CO) problem. We obtain the first global complexity result of proximal bundle method for HWC-CO. A major advantage of our analysis is that it bases on a newly defined convergence measure which is the first checkable notion of approximate  stationarity in the literature.

17:        

18: 		{\bf Key words.} hybrid weakly convex composite optimization, iteration-complexity, proximal bundle method, regularized stationary point, Moreau envelope. 

19: 		\\

20: 		

21: 		% 	{\bf Mathematics Subject Classification (2010)} 

22: 		% 	49M37 $\cdot$ 65K05 $\cdot$ 68Q25 $\cdot$ 90C25 $\cdot$ 90C30 $\cdot$ 90C60

23: 		{\bf AMS subject classifications.} 

24: 		49M37, 65K05, 68Q25, 90C25, 90C30, 90C60

25: 	\end{abstract}

26: