• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Fisheries Research >Errors in transforming length samples to age frequencies without age samples
【24h】

Errors in transforming length samples to age frequencies without age samples

机译：在没有年龄样本的情况下将长度样本转换为年龄频率的错误

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

The error characteristics of a commonly used method of transforming length samples to age frequencies when age samples are unavailable was studied by Monte Carlo simulation. Test results showed that the youngest age in the sample must be known a priori, only the younger few ages can be estimated, and that the variation of length at age is well estimated. If cohort abundance was invariant, or nearly so, a simple cutting (slicing) procedure proved best, however if cohort abundances varied fivefold or more, it failed. Unweighted least squares estimation based on proportional frequencies performed as well as or better than other methods considered. Estimation errors of the age frequency quickly increased with the variation of length at age. In the case of very low variation (CV = 0.05) the younger half of the age distribution was estimated with errors of 20% or less, but at levels of variation commonly encountered (CV approximate to 0.15) about one quarter or less of the age span was estimated with a useful level of reliability.
机译：通过蒙特卡洛模拟研究了一种在年龄样本不可用时将长度样本转换为年龄频率的常用方法的误差特性。测试结果表明，样本中最年轻的年龄必须是先验的，只能估计年龄较小的几个年龄，而且年龄的长度变化也可以得到很好的估计。如果队列的数量是不变的，或者几乎是不变的，那么简单的切割（切片）程序被证明是最好的，但是，如果队列的数量变化五倍或更多，则失败。基于比例频率的未加权最小二乘估计的执行效果优于或优于其他方法。年龄频率的估计误差随着年龄长度的变化而迅速增加。如果差异非常小（CV = 0.05），则估计年龄分布的年轻一半误差在20％或以下，但在通常遇到的差异水平（CV约为0.15）下，约为年龄的四分之一或以下估计跨度具有有用的可靠性水平。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号