Non-Parametric Logistic Regression

作者：TG. Hastie · 发表于：OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information) · 年份：2020 · 被引用次数：8 · 研究领域：Advanced Statistical Methods and Models

Linear logistic regression models the expectation of a dichotomous response variable with the model In(p(x)/( 1 -p(x))) -x' a. Often the assump tion of linearity is violated, and alternative forms are sought.In this article we deal with models of the form In.(p(x)/(l -p(x))) -C#;(zi) where the 4; are general smooth functions of the explanatory variables.Estimation is achieved using local maximum likelihood.The technique is illustrated with two examples, and is compared to existing techniques such as partial residual plots.