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

Download:

Current browse context:

stat.OT

Change to browse by:

References & Citations

Bookmark

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

Statistics > Other Statistics

Title: BFDA: A Matlab Toolbox for Bayesian Functional Data Analysis

Abstract: We provide a MATLAB toolbox, BFDA, that implements a Bayesian hierarchical model to smooth multiple functional data with the assumptions of the same underlying Gaussian process distribution, a Gaussian process prior for the mean function, and an Inverse-Wishart process prior for the covariance function. This model-based approach can borrow strength from all functional data to increase the smoothing accuracy, as well as estimate the mean-covariance functions simultaneously. An option of approximating the Bayesian inference process using cubic B-spline basis functions is integrated in BFDA, which allows for efficiently dealing with high-dimensional functional data. Examples of using BFDA in various scenarios and conducting follow-up functional regression are provided. The advantages of BFDA include: (1) Simultaneously smooths multiple functional data and estimates the mean-covariance functions in a nonparametric way; (2) flexibly deals with sparse and high-dimensional functional data with stationary and nonstationary covariance functions, and without the requirement of common observation grids; (3) provides accurately smoothed functional data for follow-up analysis.
Comments: A tool paper submitted to the Journal of Statistical Software
Subjects: Other Statistics (stat.OT); Applications (stat.AP)
Cite as: arXiv:1604.05224 [stat.OT]
  (or arXiv:1604.05224v2 [stat.OT] for this version)

Submission history

From: Jingjing Yang [view email]
[v1] Mon, 18 Apr 2016 16:06:22 GMT (907kb,D)
[v2] Fri, 3 Feb 2017 16:47:31 GMT (949kb,D)

Link back to: arXiv, form interface, contact.