1: \begin{abstract}

2: 

3: We consider the  problem of learning convolution operators associated to  compact Abelian groups.

4: We study a regularization-based approach and provide corresponding learning guarantees, discussing natural regularity condition on the convolution kernel. 

5: More precisely, we assume the convolution kernel  is  a function in a translation invariant 

6: Hilbert space and analyze a natural  ridge regression (RR) estimator. 

7: Building on existing results for RR, we  characterize the accuracy of the estimator in

8: terms of finite sample bounds. 

9: Interestingly,  regularity assumptions which are classical in the analysis of RR,  have a novel and natural interpretation  in terms of space/frequency localization. 

10: Theoretical results are illustrated by  numerical simulations.   

11: \end{abstract}

12: