1: \begin{abstract}

2: The problem of minimization of the sum of two convex functions  

3: has various theoretical and real-world applications. One of 

4: the popular methods for solving this problem is the proximal 

5: gradient method (proximal forward-backward algorithm). 

6: A very common assumption in the use of this method is that 

7: the gradient of the smooth term is globally Lipschitz continuous. However, this assumption 

8: is not always satisfied in practice, thus casting a limitation 

9: on the method. In this paper, we discuss, in a wide class of 

10: finite and infinite-dimensional spaces, a new variant 

11: of the proximal gradient method  which does not impose the 

12: above-mentioned  global Lipschitz continuity assumption. 

13: A key contribution of the method is the dependence of the iterative steps 

14: on a certain telescopic decomposition of the constraint set into subsets. 

15: Moreover, we use a Bregman divergence in the proximal 

16: forward-backward operation. Under certain practical 

17: conditions, a non-asymptotic rate of convergence (that is, 

18: in the function values) is established, as well as the weak 

19: convergence of the whole sequence to a minimizer. We also 

20: obtain a few auxiliary results of independent interest. 

21: \end{abstract}

22: