We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

stat.CO

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo

Statistics > Computation

Title: Sparse online variational Bayesian regression

Abstract: This work considers variational Bayesian inference as an inexpensive and scalable alternative to a fully Bayesian approach in the context of sparsity-promoting priors. In particular, the priors considered arise from scale mixtures of Normal distributions with a generalized inverse Gaussian mixing distribution. This includes the variational Bayesian LASSO as an inexpensive and scalable alternative to the Bayesian LASSO introduced in [65]. It also includes a family of priors which more strongly promote sparsity. For linear models the method requires only the iterative solution of deterministic least squares problems. Furthermore, for p unknown covariates the method can be implemented exactly online with a cost of $O(p^3)$ in computation and $O(p^2)$ in memory per iteration -- in other words, the cost per iteration is independent of n, and in principle infinite data can be considered. For large $p$ an approximation is able to achieve promising results for a cost of $O(p)$ per iteration, in both computation and memory. Strategies for hyper-parameter tuning are also considered. The method is implemented for real and simulated data. It is shown that the performance in terms of variable selection and uncertainty quantification of the variational Bayesian LASSO can be comparable to the Bayesian LASSO for problems which are tractable with that method, and for a fraction of the cost. The present method comfortably handles $n = 65536$, $p = 131073$ on a laptop in less than 30 minutes, and $n = 10^5$, $p = 2.1 \times 10^6$ overnight.
Subjects: Computation (stat.CO); Numerical Analysis (math.NA); Optimization and Control (math.OC); Machine Learning (stat.ML)
Journal reference: SIAM/ASA Journal on Uncertainty Quantification 10.3 (2022): 1070-1100
DOI: 10.1137/21M1401188
Cite as: arXiv:2102.12261 [stat.CO]
  (or arXiv:2102.12261v2 [stat.CO] for this version)

Submission history

From: Kody Law [view email]
[v1] Wed, 24 Feb 2021 12:49:42 GMT (2702kb)
[v2] Wed, 22 Dec 2021 15:36:47 GMT (2268kb,D)

Link back to: arXiv, form interface, contact.