摘要
用几种不同的密度推定法,对测线调查样本进行密度推定。使用的最佳带幅宽有的依赖于探知函数,有的不依赖探知函数,还用到经典的最大似然估计法。把5种密度推定法用于计算机模拟产生测线调查的距离数据,由统计量RRMSE和RB对推定值进行评价,来验证其优劣。
In this paper, we use several kinds of density estimate, carry on density estimate to line transect sampling of computer simulation. Some of optimal bandwidths used depends on detection function, some do not rely on detection function, still use classic ground Maximum Likelihood estimate. We produce distance data of the line transect to imitate with 5 density estimations, RRMSE and RB are used to appraise the estimate value and verify their quality.
出处
《江西教育学院学报》
2004年第6期6-9,共4页
Journal of Jiangxi Institute of Education
关键词
调查
生物分布
依赖
样本
密度
评价
验证
推定
产生
line transect sampling
detection function
smoothing
relative root mean square error
relative bias
Least Squares Cross Validation