科研之友
微信
新浪微博
Facebook
分享链接
摘要
In this paper, we study the classical nonconvex linearly constrained optimization problem. Under some mild conditions, we obtain that the penalization sequence is nonincreasing and the sequence generated by our algorithm has finite length. Based on the assumption that the objective functions have Kurdyka-Lojasiewicz property, we prove the convergence of the algorithm. We also show the numerical efficiency of our method by the concrete applications in the areas of image processing and statistics.
- 出版日期2015-8
- 单位重庆大学
