TY - JOUR
TI - Optimization of ℓp-regularized Linear Models via Coordinate Descent
AU - Klimaszewski, Jacek
AU - Korzeń, Marcin
TI - Optimization of ℓp-regularized Linear Models via Coordinate Descent
AB - 
	In this paper we demonstrate, how `p-regularized univariate quadratic loss function can be effectively optimized (for 0 6 p 6 1) without approximation of penalty term and provide analytical solution for p = 1 2 . Next we adapt this approach for important multivariate cases like linear and logistic regressions, using Coordinate Descent algorithm. At the end we compare sample complexity of `1 with `p, 0 6 p < 1 regularized models for artificial and real datasets.
VL - 2016
IS - Volume 25
PY - 2017
SN - 1732-3916
C1 - 2083-8476
SP - 61
EP - 72
DO - 10.4467/20838476SI.16.005.6186
UR - https://ejournals.eu/en/journal/schedae-informaticae/article/optimization-of-p-regularized-linear-models-via-coordinate-descent
KW - Classification
KW - Coordinate Descent
KW - Regression
KW - Sparsity