FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W1882912287> ?p ?o ?g. }

• W1882912287 abstract "Variable selection and model choice are of major concern in many statisticalapplications, especially in regression models for high-dimensional data.Boosting is a convenient statistical method that combines model fitting withintrinsic model selection. We investigate the impact of base-learnerspecification on the performance of boosting as a model selection procedure. Weshow that variable selection may be biased if the base-learners have differentdegrees of flexibility, both for categorical covariates and for smooth effectsof continuous covariates. We investigate these problems from a theoreticalperspective and suggest a framework for unbiased model selection based on ageneral class of penalized least squares base-learners. Making all base-learnerscomparable in terms of their degrees of freedom strongly reduces the selectionbias observed with naive boosting specifications. Furthermore, the definition ofdegrees of freedom that is used in the smoothing literature is questionable inthe context of boosting, and an alternative definition is theoretically derived.The importance of unbiased model selection is demonstrated in simulations and inan application to forest health models.A second aspect of this thesis is the expansion of the boosting algorithm to newestimation problems: by using constraint base-learners, monotonicity constrainedeffect estimates can be seamlessly incorporated in the existing boostingframework. This holds for both, smooth effects and ordinal variables.Furthermore, cyclic restrictions can be integrated in the model for smootheffects of continuous covariates. In particular in time-series models, cyclicconstraints play an important role. Monotonic and cyclic constraints of smootheffects can, in addition, be extended to smooth, bivariate function estimates.If the true effects are monotonic or cyclic, simulation studies show thatconstrained estimates are superior to unconstrained estimates. In three casestudies (the modeling the presence of Red Kite in Bavaria, the modeling ofactivity profiles for Roe Deer, and the modeling of deaths caused by airpollution in Sao Paulo) it is shown that both constraints can be integratedin the boosting framework and that they are easy to use.All described results were included in the R add-on package mboost." @default.
• W1882912287 created "2016-06-24" @default.
• W1882912287 creator A5072285340 @default.
• W1882912287 date "2011-12-05" @default.
• W1882912287 modified "2023-10-02" @default.
• W1882912287 title "Boosting in structured additive models" @default.
• W1882912287 cites W1487825358 @default.
• W1882912287 cites W1491862040 @default.
• W1882912287 cites W1528145197 @default.
• W1882912287 cites W1535538352 @default.
• W1882912287 cites W1554944419 @default.
• W1882912287 cites W1556631438 @default.
• W1882912287 cites W1584023832 @default.
• W1882912287 cites W1598401317 @default.
• W1882912287 cites W1602479663 @default.
• W1882912287 cites W163215356 @default.
• W1882912287 cites W1663521020 @default.
• W1882912287 cites W1678356000 @default.
• W1882912287 cites W182248535 @default.
• W1882912287 cites W1949314668 @default.
• W1882912287 cites W1965267599 @default.
• W1882912287 cites W1965489797 @default.
• W1882912287 cites W1965755064 @default.
• W1882912287 cites W1970274193 @default.
• W1882912287 cites W1975075722 @default.
• W1882912287 cites W1975354406 @default.
• W1882912287 cites W1982542505 @default.
• W1882912287 cites W1982763475 @default.
• W1882912287 cites W1983440945 @default.
• W1882912287 cites W1988790447 @default.
• W1882912287 cites W1991047917 @default.
• W1882912287 cites W1992549534 @default.
• W1882912287 cites W2002146930 @default.
• W1882912287 cites W200887292 @default.
• W1882912287 cites W2008950066 @default.
• W1882912287 cites W2012205546 @default.
• W1882912287 cites W2013182691 @default.
• W1882912287 cites W2024046085 @default.
• W1882912287 cites W2025266808 @default.
• W1882912287 cites W2026257972 @default.
• W1882912287 cites W2027106381 @default.
• W1882912287 cites W2034433729 @default.
• W1882912287 cites W2039380015 @default.
• W1882912287 cites W2039656279 @default.
• W1882912287 cites W2042695116 @default.
• W1882912287 cites W2045495405 @default.
• W1882912287 cites W2053934160 @default.
• W1882912287 cites W2062445441 @default.
• W1882912287 cites W2067800992 @default.
• W1882912287 cites W2073587703 @default.
• W1882912287 cites W2076448121 @default.
• W1882912287 cites W2088883866 @default.
• W1882912287 cites W2091041003 @default.
• W1882912287 cites W2093717447 @default.
• W1882912287 cites W2097601813 @default.
• W1882912287 cites W2099372171 @default.
• W1882912287 cites W2102201073 @default.
• W1882912287 cites W2109785413 @default.
• W1882912287 cites W2112076978 @default.
• W1882912287 cites W2112081648 @default.
• W1882912287 cites W2114139895 @default.
• W1882912287 cites W2115103545 @default.
• W1882912287 cites W2117897510 @default.
• W1882912287 cites W2121203842 @default.
• W1882912287 cites W2122291037 @default.
• W1882912287 cites W2123998733 @default.
• W1882912287 cites W2127441697 @default.
• W1882912287 cites W2131793865 @default.
• W1882912287 cites W2135993243 @default.
• W1882912287 cites W2139857790 @default.
• W1882912287 cites W2140308441 @default.
• W1882912287 cites W2146766088 @default.
• W1882912287 cites W2154065358 @default.
• W1882912287 cites W2156909104 @default.
• W1882912287 cites W2156963621 @default.
• W1882912287 cites W2167834874 @default.
• W1882912287 cites W2168464387 @default.
• W1882912287 cites W2171793733 @default.
• W1882912287 cites W2201025417 @default.
• W1882912287 cites W2315221623 @default.
• W1882912287 cites W2319794630 @default.
• W1882912287 cites W2369090271 @default.
• W1882912287 cites W2479395228 @default.
• W1882912287 cites W2500597054 @default.
• W1882912287 cites W2765396776 @default.
• W1882912287 cites W2797583072 @default.
• W1882912287 cites W2797925981 @default.
• W1882912287 cites W2911964244 @default.
• W1882912287 cites W2912273251 @default.
• W1882912287 cites W3020967475 @default.
• W1882912287 cites W3099692195 @default.
• W1882912287 cites W627957585 @default.
• W1882912287 hasPublicationYear "2011" @default.
• W1882912287 type Work @default.
• W1882912287 sameAs 1882912287 @default.
• W1882912287 citedByCount "6" @default.
• W1882912287 countsByYear W18829122872012 @default.
• W1882912287 countsByYear W18829122872014 @default.
• W1882912287 countsByYear W18829122872016 @default.
• W1882912287 countsByYear W18829122872017 @default.

• next