Output details
11 - Computer Science and Informatics
University of Edinburgh
Return to search
Output 0 of 0 in the submission
Output title
Elliptical slice sampling
Type
E - Conference contribution
DOI
-
Name of conference/published proceedings
Proceedings of the 13th International Conference on Artificial Intelligence and Statistics (AISTATS)
Volume number
9
Issue number
-
First page of article
541
ISSN of proceedings
-
Year of publication
2010
Number of additional authors
2
Additional information
<24> Originality: First effective Monte Carlo method for latent Gaussian models with zero free parameters, provided online in one simple function. Contemporary work (Titsias, Lawrence, Rattray, NIPS 2009) came with complex software packages.
Significance: The work has allowed rapid development of codes for the authors' and others' nonparametric Bayesian models. A companion paper "Slice sampling covariance hyperparameters of latent Gaussian models" was awarded an oral at NIPS (2010, 2% submissions).
Rigour: Generalized Wishart Processes (Wilson and Ghahramani, best student paper award UAI 2011) reports it works as advertised, and "we have found Elliptical Slice Sampling incredibly robust".
Interdisciplinary
-
Cross-referral requested
-
Research group
B - Institute for Adaptive & Neural Computation
Citation count
12
Proposed double-weighted
No
Double-weighted statement
-
Reserve for a double-weighted output
No
Non-English
No
English abstract
-
