In hofnerb/mboost: Model-Based Boosting

Scholary papers describing the methodology

library("RefManageR")
bib <- ReadBib("TH.bib", check = FALSE)
BibOptions(check.entries = FALSE, style = "html", cite.style = "authoryear",
           bib.style = "numeric")

The mboost package implements componentwise functional gradient descent boosting, originally invented by r Citet(bib, "Buehlmann_Yu_2003"). The theory and package are described in the review by r Citet(bib, "Buehlmann:2008:StatSci"), an overview about the implementation is given in r Citet(bib, "Hothorn+Buehlmann+Kneib+Schmid+Hofner_2010") and r Citet(bib, "Hothorn:2006:Bioinformatics:16940323"). A tutorial can be found in r Citet(bib, "Hofner_Mayr_Robinzonov_2014").

Applications in survival analysis are described in r Citet(bib, "Hothorn:2006:Biostatistics:16344280"), r Citet(bib, "Schmid+Hothorn:2008a") and r Citet(bib, "Mayr_Schmid_2014").

Geoadditive models have been dealt with by r Citet(bib, "Kneib+Hothorn+Tutz:2009"), r Citet(bib, "Hothorn+Mueller+Schroeder_2011") and r Citet(bib, "Schmid_Hothorn_Maloney_Weller_Potapov_2011"), and, with special emphasis on monotonicity constraints, by r Citet(bib, "Hofner_Mueller_Hothorn_2011").

Boosted quantile regression models were introduced by r Citet(bib, "Fenske+Kneib+Hothorn_2011") and later extended to longitudinal data by r Citet(bib, "Fenske_Fahrmeir_Hothorn_2013") and to the derivation of prediction intervals by r Citet(bib, "Mayr_Hothorn_Fenske_2012").

The application of P-spline base-learners was discussed by r Citet(bib, "Schmid+Hothorn:2008b"). Unbiased model selection was implemented as described by r Citet(bib, "Hofner+Hothorn+Kneib_2011"). An approach to non-linear time series can be found in r Citet(bib, "Robinzonov_Tutz_Hothorn_2012"). Boosted classifiers based on a direct optimisation of the partial AUC were introduced by r Citet(bib, "Schmid_Hothorn_Krause_2012").

r Citet(bib, "Hothorn_Kneib_Buehlmann_2014") used componentwise array boosting to fit a novel class of conditional transformation models; some simplifications are given by r Citet(bib, "Moest_Schmid_Faschingbauer_2014"). In a similar spirit, r Citet(bib, "Mayr_Fenske_Hofner_2012") use the package to fit GAMLSS models.

References

PrintBibliography(bib, .opts = list(check.entries = FALSE, sorting = "ynt"))

hofnerb/mboost documentation built on Jan. 10, 2024, 9:21 p.m.