swMATH ID: 14769
Software Authors: Jingjing Yang, Peng Ren
Description: BFDA: A Matlab Toolbox for Bayesian Functional Data Analysis. We provide a Matlab toolbox, BFDA, that implements a Bayesian hierarchical model for smoothing functional data and estimating mean-covariance functions simultaneously and nonparametricaly, with the assumptions of Gaussian process for functional data and mean function, and the assumption of Inverse-Whishart process for the covariance function. An option of approximating the Bayesian inference process with cubic B-spline basis functions is integrated in this toolbox, which allows the possibility of dealing with large-scale functional data. Examples of functional data regression with one functional independent variable, scalar and functional response variables are provided. The advantages of BFDA include: (1) Simultaneously smooths functional data and estimates the mean-covariance functions in a nonparametric way; (2) efficiently deals with large-scale functional data with random or high-dimensional observation-grids; (3) Flexibly adapts for both stationary and nonstationary functional data; (4) Provides accurately smoothed functional data for follow-up analysis.
Homepage: http://arxiv.org/abs/1604.05224
Dependencies: Matlab
Keywords: FDA; functional data analysis; Bayesian hierarchical model; Gaussian process; cubic B-spline basis functions; MATLAB; Journal of Statistical Software; jstatsoft.org
Related Software: GPFDA; Matlab; fda (R); fda.usc; devtools; ggplot2; pbapply; funHDDC; caret; refund; neuralnet; FDboost; funFEM; RSNNS; future; keras; R; FuncNN; PACE; MCMC Diagnostics
Cited in: 0 Publications