FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2303763562> ?p ?o ?g. }

• W2303763562 abstract "Many studies produce dependent data. For example, repeated measures experiments produce repeated measurements on independent subjects and longitudinal studies generate datasets with several observations collected over time. These data have a clustered correlated structure. The clusters are defined as groups of observations that are correlated, but are uncorrelated with observations in other clusters. Generalized Estimating Equations (GEEs) (Zeger and Liang, 1986) are estimating functions used to estimate mean models for cluster correlated data when only the mean model and the variance function (of the mean) have been specified. They are often appropriate for modeling the population mean of counts, binary, and continuous responses when the correlation structure is a nuisance rather than the focus. The variance function of the mean can be taken from a distribution in the exponential family, though it need not be, since GEE makes no distributional assumptions beyond the first two moments. Like many other statistical models, the GEE mean model and variance function may be misspecified. Robustification against variance function misspecification, especially correlation structure misspecification, is done through a semiparametric “sandwich” variance estimation of the standard errors. Robustification of the mean function is possible through a kernel-based localization of GEEs, which makes weak assumptions about the functional form of the mean function. The work presented here concerns a situation that lies in-between the two extremes of knowing the mean model entirely and not knowing it at all. We imagine that a model builder specifies a parametric model that he or she believes may only be partly correct. Then hybridization of global (parametric) and local (nonparametric) models is used to formulate a semiparametric mean model that draws on the strengths of global and local regression for estimation in this setting. We seek then to robustify the GEE mean model specification using local GEEs. The ideas presented here are an extension of the works on model robust regression (MRR) by Mays, Birch, and Starnes (2001) for the linear model, and model robust regression for quantal regression by Nottingham and Birch (2000). This article is organized as follows. In section 2 we review parametric, nonparametric and semiparametric GEEs and discuss bandwidth selection for local GEEs. In section 3 we review MRR, propose extensions of MRR to GEEs, and give cross-validation and optimal mixing parameter estimators. An example is presented in section 4. A discussion follows in section 5." @default.
• W2303763562 created "2016-06-24" @default.
• W2303763562 creator A5003244886 @default.
• W2303763562 date "2002-03-29" @default.
• W2303763562 modified "2023-09-26" @default.
• W2303763562 title "Model Robust Regression Based on Generalized Estimating Equations" @default.
• W2303763562 cites W135795571 @default.
• W2303763562 cites W1440417091 @default.
• W2303763562 cites W1484648503 @default.
• W2303763562 cites W1490266132 @default.
• W2303763562 cites W1493207670 @default.
• W2303763562 cites W1528905581 @default.
• W2303763562 cites W1576686186 @default.
• W2303763562 cites W1597926580 @default.
• W2303763562 cites W1598266570 @default.
• W2303763562 cites W1668720348 @default.
• W2303763562 cites W1963509555 @default.
• W2303763562 cites W1967320885 @default.
• W2303763562 cites W1969577722 @default.
• W2303763562 cites W1971713783 @default.
• W2303763562 cites W1975285668 @default.
• W2303763562 cites W1976429900 @default.
• W2303763562 cites W1976834773 @default.
• W2303763562 cites W1978136073 @default.
• W2303763562 cites W1978930761 @default.
• W2303763562 cites W1979159029 @default.
• W2303763562 cites W1980316423 @default.
• W2303763562 cites W1984140225 @default.
• W2303763562 cites W1985875898 @default.
• W2303763562 cites W1991158845 @default.
• W2303763562 cites W1997927050 @default.
• W2303763562 cites W2000098294 @default.
• W2303763562 cites W2002374079 @default.
• W2303763562 cites W2005843344 @default.
• W2303763562 cites W2008520227 @default.
• W2303763562 cites W2014268383 @default.
• W2303763562 cites W2024081693 @default.
• W2303763562 cites W2024332750 @default.
• W2303763562 cites W2024983310 @default.
• W2303763562 cites W2027244596 @default.
• W2303763562 cites W2027248710 @default.
• W2303763562 cites W2032320709 @default.
• W2303763562 cites W2035240057 @default.
• W2303763562 cites W2037489600 @default.
• W2303763562 cites W2037767836 @default.
• W2303763562 cites W2042498294 @default.
• W2303763562 cites W2045606408 @default.
• W2303763562 cites W2046977371 @default.
• W2303763562 cites W2050297026 @default.
• W2303763562 cites W2052128863 @default.
• W2303763562 cites W2053068905 @default.
• W2303763562 cites W2057032881 @default.
• W2303763562 cites W2059507684 @default.
• W2303763562 cites W2064385888 @default.
• W2303763562 cites W2065890263 @default.
• W2303763562 cites W2067192762 @default.
• W2303763562 cites W2069888879 @default.
• W2303763562 cites W2072518584 @default.
• W2303763562 cites W2072698722 @default.
• W2303763562 cites W2073208705 @default.
• W2303763562 cites W2073530080 @default.
• W2303763562 cites W2074673068 @default.
• W2303763562 cites W2078658856 @default.
• W2303763562 cites W2083610611 @default.
• W2303763562 cites W2086067392 @default.
• W2303763562 cites W2086403511 @default.
• W2303763562 cites W2091886411 @default.
• W2303763562 cites W2093743402 @default.
• W2303763562 cites W2094670798 @default.
• W2303763562 cites W2099726211 @default.
• W2303763562 cites W2101424588 @default.
• W2303763562 cites W2112876018 @default.
• W2303763562 cites W2118091795 @default.
• W2303763562 cites W2118381969 @default.
• W2303763562 cites W2129905273 @default.
• W2303763562 cites W2130691021 @default.
• W2303763562 cites W2132890483 @default.
• W2303763562 cites W2136987861 @default.
• W2303763562 cites W2137295153 @default.
• W2303763562 cites W2145816995 @default.
• W2303763562 cites W2146703155 @default.
• W2303763562 cites W2149860264 @default.
• W2303763562 cites W2150105594 @default.
• W2303763562 cites W2164293083 @default.
• W2303763562 cites W2166163519 @default.
• W2303763562 cites W2170821252 @default.
• W2303763562 cites W2211925278 @default.
• W2303763562 cites W2260664148 @default.
• W2303763562 cites W2315786704 @default.
• W2303763562 cites W2319794630 @default.
• W2303763562 cites W2472024274 @default.
• W2303763562 cites W2556205749 @default.
• W2303763562 cites W2580802234 @default.
• W2303763562 cites W2582653597 @default.
• W2303763562 cites W2587346059 @default.
• W2303763562 cites W2724315023 @default.
• W2303763562 cites W27605680 @default.
• W2303763562 cites W2796930163 @default.
• W2303763562 cites W2797333853 @default.
• W2303763562 cites W2797458901 @default.

• next