1: \begin{abstract}

2: 	

3: We present a novel scheme to accurately predict atomic forces as vector

4: quantities, rather than sets of scalar components, by Gaussian Process

5: (GP) Regression. This is based on matrix-valued kernel functions, on

6: which we impose the requirements that the predicted force rotates with the target configuration

7: and is independent of any rotations applied to the configuration database

8: entries. We show that such covariant GP kernels can be obtained

9: by integration over the elements of the rotation group $SO(d)$ for

10: the relevant dimensionality $d$. 

11: %

12: Remarkably, in specific cases the

13: integration can be carried out analytically and yields a conservative

14: force field that can be recast into a pair interaction form. 

15: %

16: Finally,

17: we show that restricting the integration to a summation over the elements

18: of a finite point group relevant to the target system is sufficient

19: to recover an accurate GP. The accuracy of our kernels in predicting

20: quantum-mechanical forces in real materials is investigated by tests

21: on pure and defective Ni, Fe and Si crystalline systems.

22: \end{abstract}

23: