oscar

Optimal Subset Cardinality Regression (OSCAR) Models Using the L0-Pseudonorm

Teemu Daniel Laajala, Anni Halkola, Kaisa Joki

Last update: 2 October 2023

Copyright license: GNU General Public License, version 3.0

Software version identifier: 1.1.4, 1.0.3, 1.0.4, 1.1.2-2, 1.1.7, 1.2.0, 1.2.1

Description

Optimal Subset Cardinality Regression (OSCAR) models offer regularized linear regression using the L0-pseudonorm, conventionally known as the number of non-zero coefficients. The package estimates an optimal subset of features using the L0-penalization via cross-validation, bootstrapping and visual diagnostics. Effective Fortran implementations are offered along the package for finding optima for the DC-decomposition, which is used for transforming the discrete L0-regularized optimization problem into a continuous non-convex optimization task. These optimization modules include DBDC ('Double Bundle method for nonsmooth DC optimization' as described in Joki et al. (2018) <doi:10.1137/16M1115733>) and LMBM ('Limited Memory Bundle Method for large-scale nonsmooth optimization' as in Haarala et al. (2004) <doi:10.1080/10556780410001689225>). The OSCAR models are comprehensively exemplified in Halkola et al. (2023) <doi:10.1371/journal.pcbi.1010333>). Multiple regression model families are supported: Cox, logistic, and Gaussian.

Cites work

This page was built for software: oscar