计算机工程与应用 ›› 2012, Vol. 48 ›› Issue (6): 38-41.

• 研究、探讨 • 上一篇    下一篇

利用矢量基学习和自适应迭代算法改进LSSVR

王鲜芳1,杜志勇2   

  1. 1.河南师范大学 计算机与信息技术学院,河南 新乡 453007
    2.河南机电高等专科学校,河南 新乡 453007
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2012-02-21 发布日期:2012-02-21

Using vector-base learning and adaptive iterative algorithm to improve LSSVR

WANG Xianfang1, DU Zhiyong2   

  1. 1.School of Computer and Information Technology, Henan Normal University, Xinxiang, Henan 453007, China
    2.Henan Mechanical and Electrical Engineering College, Xinxiang, Henan 453007, China
  • Received:1900-01-01 Revised:1900-01-01 Online:2012-02-21 Published:2012-02-21

摘要: 针对最小二乘支持向量回归缺乏传统SVR的稀疏性和鲁棒性等问题,综合矢量基学习和自适应迭代算法的优势,提出了一种改进的加权最小二乘支持向量回归算法(LSSVR)。该算法通过引入用矢量基学习和自适应迭代相结合的方式得到一个小的支持向量集,可以避免递推时可能出现的误差积累问题,有效提高算法的稀疏性和稳定性;同时采用加权方法确定权值系数以减小训练样本中非高斯噪声的影响。实验结果表明,改进的LSSVR具有较好的鲁棒性、支持向量稀疏性和动态建模实时性。

关键词: 矢量基, 自适应迭代算法, 支持向量稀疏性

Abstract: Combining the advantages of the vector-based learning and adaptive iterative algorithm, an improved weighted Least Squares Support Vector Regression(LSSVR) is proposed to solve the problems of the least squares support vector regression methods, such as lacking of sparsely and robustly. During the training process of algorithm, the vector-base learning and automatic iterative procedures are introduced and a small support vector set can be obtained adaptively. This method can avoid the error accumulation during the iterative processing and improve the sparseness and stability of the algorithm, while the weights are determined by a robust method in order to reduce the effect of the outliers(e.g.resulting from non-Gaussian noise). The experimental results show that the proposed algorithm has a better robust, sparsely of support vector and real-time performance of dynamic modeling.

Key words: vector-base, adaptive iterative algorithm, support vector sparsely