Bayesian sparse convex clustering via global-local shrinkage priors
| Autor: | Kaito Shimamura, Shuichi Kawano |
|---|---|
| Rok vydání: | 2021 |
| Předmět: |
FOS: Computer and information sciences
Statistics and Probability Computer Science - Machine Learning Computer science Bayesian probability Machine Learning (stat.ML) Feature selection Computer Science::Digital Libraries Regularization (mathematics) Machine Learning (cs.LG) Methodology (stat.ME) Statistics::Machine Learning symbols.namesake Statistics - Machine Learning Prior probability Cluster analysis Statistics - Methodology Markov chain Monte Carlo Computational Mathematics ComputingMethodologies_PATTERNRECOGNITION Norm (mathematics) Computer Science::Mathematical Software symbols Statistics Probability and Uncertainty Algorithm Gibbs sampling |
| Zdroj: | Computational Statistics. 36:2671-2699 |
| ISSN: | 1613-9658 0943-4062 |
| DOI: | 10.1007/s00180-021-01101-7 |
| Popis: | Sparse convex clustering is to group observations and conduct variable selection simultaneously in the framework of convex clustering. Although a weighted $$L_1$$ L 1 norm is usually employed for the regularization term in sparse convex clustering, its use increases the dependence on the data and reduces the estimation accuracy if the sample size is not sufficient. To tackle these problems, this paper proposes a Bayesian sparse convex clustering method based on the ideas of Bayesian lasso and global-local shrinkage priors. We introduce Gibbs sampling algorithms for our method using scale mixtures of normal distributions. The effectiveness of the proposed methods is shown in simulation studies and a real data analysis. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full text from SpringerLink