Abstract:
Aiming at the problem of the parameter estimation of the binary probit regression models, this paper proposes a novel algorithm to estimate parameter based on Particle Swarm Optimization(PSO) algorithm. Maximum likelihood estimation rule is adopted to be fitness function for the PSO algorithm. The model of computing parameter to the binary probit regression model is set up. Through a numerical simulation computational experiment, the effectiveness of this algorithm is demonstrated for the parameter estimation problem of the binary probit regression models. Numerical simulation analysis shows that the algorithm has better goodness of fittest in the regression result.
Key words:
Prticle Swarm Optimization(PSO) algorithm,
parameter estimation,
binary probit regression model,
maximum likelihood estimation
摘要: 针对二值probit回归模型中的参数估计问题,提出一种基于粒子群优化(PSO)的参数估计算法。该算法采用以最大似然准则作为PSO的适应度函数,建立二值probit回归模型中的参数估计计算模型。数值仿真分析表明,该算法性能较好,回归结果具有较高的拟合 优度。
关键词:
粒子群优化算法,
参数估计,
二值probit回归模型,
最大似然估计
CLC Number:
LIU Jin-ping; ; YU Jin-xiang. Parameter Estimation of probit Model Based on Particle Swarm Optimization Algorithm[J]. Computer Engineering, 2009, 35(23): 198-200.
刘锦萍;郁金祥. 基于PSO算法的probit模型参数估计[J]. 计算机工程, 2009, 35(23): 198-200.