1: \begin{abstract}

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

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

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

5: 		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.}

6: 	

7: \noindent{\bf Keywords:} 

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

9: \end{abstract}

10: