1: \begin{abstract}

2: % Body of abstract:

3: { In this paper,  we focus on  the local convergence rate analysis of the proximal iteratively reweighted $\ell_1$ algorithms for solving $\ell_p$ regularization

4:  problems, which are widely applied for inducing sparse solutions. 

5: We show that if the  \emph{Kurdyka-\L ojasiewicz} (KL)  property  is satisfied,  the algorithm converges 

6: to  a unique first-order stationary  point; furthermore, the algorithm  has local  linear convergence   or local sublinear convergence. The theoretical results we derived are much stronger than the existing results for iteratively reweighted $\ell_1$ algorithms.}

7: % Keywords:

8:   Kurdyka-\L ojasiewicz property, $\ell_p$ regularization, iteratively reweighted algorithm 

9: \end{abstract}

10: