Mathematical Sciences: Transformation Techniques in Regression

本研究的重点是回归中的转换 分析时，有一个理论模型的中位数 响应，但误差分布必须根据经验建模。 转换的目的是稳定方差 并引入对称分布的误差。 估计这样 理论模型中的参数估计。 它也可以 帮助模拟物理系统的随机性， 数据出现了。 这项统计科学的研究是在 回归分析的一般领域。 数学理论在这方面 区域处理测量之间的关联程度 变量和评估这种关联的性质。 这 工作扩展了可用的数学模型的灵活性 进行这种评估，从而将允许更丰富的多样性 我们需要去发现和描述这些关系。

This research focuses on transformations in regression analysis when there is a theoretical model for the median response, but the error distribution must be modeled empirically. The purposes of the transformation are to stabilize the variance and to induce symmetrically distributed errors. Estimating such timation of parameters in the theoretical model. It can also help model the random nature of the physical system from which the data arose. This reseach in statistical science is in the general area of regression analysis. Mathematical theory in this area deals with measuring the degree of association between variables and assessing the nature of such associations. This work extends the flexibility of the mathematical models available for making such assessments and thus will permit a richer variety of relationships to be discovered and characterized.