1: \begin{abstract}

2: We consider linear ill-conditioned operator equations in a Hilbert space setting. 

3: Motivated by the aggregation method, we consider approximate solutions constructed  

4: from linear combinations of Tikhonov regularization, which amounts to finding 

5: solutions in a rational Krylov space. By mixing these with usual Krylov spaces, we

6: consider least-squares problem in these mixed rational  spaces. Applying the Arnoldi method leads

7: to a sparse, pentadiagonal representation of the forward operator, and we introduce 

8: the Lanczos method for solving the least-squares problem by factorizing this matrix. 

9: Finally, we present an equivalent conjugate-gradient-type method that does not rely 

10: on explicit orthogonalization but uses short-term recursions and Tikhonov regularization

11: in each second step. We illustrate the convergence and regularization properties by some numerical examples.

12: \end{abstract}