欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
传统的线性规划支持向量回归算法需要选择一个折中参数C来确定经验风险和置信风险之间的比例,而针对不同的数据选择最优的参数C一般并不容易。为解决这一问题,提出一种给定经验风险水平的线性规划支持向量回归算法,该算法能够事先确定经验风险水平的大小。另外,新算法还可以通过设置不同样本点上经验风险的大小,处理样本中存在异方差的情况。仿真试验验证了所给算法的可行性和有效性。
英文摘要：
Traditional linear programming support vector regression algorithm needs to choose a tradeoff coefficient C for making certain the proportion between the empirical risk and the confidence risk, generally speaking it is not easy to choose an optimal C in correspondence to different data. In order to solving the problem, a linear programming support vector regression with given empirical risk is proposed, it can make sure the extent of empirical risk in advance. In addition, the new algorithm can also solve the problem of heterogeneity of variance existing in samples by setting different empirical risks for different samples. The results of simulative experiments verify the feasibility and effectiveness of the proposed algorithm.
同期刊论文项目
同项目期刊论文
Approximation quality of fuzzy preference rough sets and its application in radio signal recognition
Properties of fuzzy induced ordered weighted averaging aggregation operator and its application in d
Implication Operator on the Set of ∨-Irreducible Element in the Linguistic Truth-Valued Intuitionist
期刊信息
位置：