家畜生产中生存模型的分析方法及比较研究

发布时间：2018-02-22 03:35

本文关键词： 生存分析模型截尾数据家畜生产　出处：《畜牧与兽医》2017年12期 　论文类型：期刊论文

【摘要】：本文介绍了生存分析的概念和特点。对线性模型、随机回归模型和比例风险模型等多种生存模型在家畜生产中的应用进行了比较,重点讨论了比例风险模型的分布、非参数估计以及该模型的扩展应用。结果表明:线性模型适合把生存性状定义成二元性状进行分析,但局限是很难解释截尾问题;随机回归模型可作为利用线性模型来分析截尾数据,但当数据量较少或缺失时,很难处理时间依赖型变量;比例风险模型尽管目前尚不能在超过两个性状的多元分析中得到应用,但因其比较适合处理生存数据,特别是分析截尾数据或时间依赖型变量,在动物育种数据分析时有广泛应用。
[Abstract]:This paper introduces the concept and characteristics of survival analysis, compares the applications of several survival models, such as linear model, stochastic regression model and proportional risk model, in livestock production, and discusses the distribution of proportional risk model. The nonparametric estimation and the extended application of the model show that the linear model is suitable for the analysis of survival traits as binary characters, but the limitation is that it is difficult to explain the truncation problem. The stochastic regression model can be used as a linear model to analyze truncated data, but it is difficult to deal with time-dependent variables when the amount of data is less or missing. Although the proportional risk model can not be used in multivariate analysis of more than two traits at present, it is more suitable for dealing with survival data, especially for truncated data or time-dependent variables. It is widely used in animal breeding data analysis.
【作者单位】：海南大学农学院;华南农业大学动物科学学院;
【基金】：2016年国家级大学生创新创业训练计划项目(201610589019)
【分类号】：S82

【相似文献】