Paper Info
Title
Coordinate Descent Algorithm for Normal-Likelihood-Based Group Lasso in Multivariate Linear Regression
Abstract
We focus on an optimization algorithm for a normal-likelihood-based group Lasso in multivariate linear regression. A negative multivariate normal log-likelihood function with a block-norm penalty is used as the objective function. A solution for the minimization problem of a quadratic form with a norm penalty is given without using the Karush-Kuhn-Tucker condition. In special cases, the minimization problem can be solved without solving simultaneous equations of the first derivatives. We derive update equations of a coordinate descent algorithm for minimizing the objective function. Further, by using the result of the special case, we also derive update equations of an iterative thresholding algorithm for minimizing the objective function.
Year
DOI
Venue
2021
10.1007/978-981-16-2765-1_36
INTELLIGENT DECISION TECHNOLOGIES, KES-IDT 2021
Keywords
DocType
Volume
Adaptive group Lasso, Block-norm regularization, Multivariate linear regression, Negative normal log-likelihood function
Conference
238
ISSN
Citations 
PageRank 
2190-3018
0
0.34
References 
Authors
0
2
Name
Order
Citations
PageRank
Hirokazu Yanagihara1218.66
Ryoya Oda200.34