University of Limerick
Browse
Mackenzie_2011_LASSO.pdf (938.84 kB)

LASSO penalised likelihood in high-dimensional contingency tables

Download (938.84 kB)
conference contribution
posted on 2022-11-11, 19:13 authored by Susana Conde, Gilbert MackenzieGilbert Mackenzie
We consider several least absolute shrinkage and selection operator (LASSO) penalized likelihood approaches in high dimensional contingency tables and with hierarchical log-linear models. These include the proposal of a parametric, analytic, convex, approximation to the LASSO. We compare them with "classical" stepwise search algorithms. The results show that both backwards elimination and forward selection algorithms select more parsimonious (i.e. sparser) models which are always hierarchical, unlike the competing LASSO techniques.

History

Publication

Proceedings of the 26th International Workshop on Statistical Modelling;

Publisher

IWSM

Note

peer-reviewed

Other Funding information

Glaxosmithkline (GSK), SFI

Language

English

Also affiliated with

  • BIO-SI - Bio-Statistics & Informatics Project

Department or School

  • Mathematics & Statistics

Usage metrics

    University of Limerick

    Categories

    No categories selected

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC