File: readme
Date: January 8, 2008
Author: Jo-Anne Ting
Matlab source code:
run_parameterEst_test.m : Test file to run. Creates synthetic data
sets and compares the performances of
algorithms on a parameter estimation
problem for noisy high-dimensional linear
regression.
noisyBLR.m : Source code for Bayesian de-noising
regression algorithm. Estimates the
regression coefficient given
noise-contaminated input and output data
stepwise_regression.m : Wrapper interface function for Matlab's
stepwisefit function
arrfit.m : Source code for Adaptive Ridge Regression
pls_v1.m : Source code for PLS regression
fapbf.m : Source code for Joint-space version of
factor analysis for regression
Comments:
Run run_parameterEst_test.m. The indicator variables can be
toggled on or off, depending on which algorithms the user wants to run.
Please refer to Ting, D'Souza & Schaal (2009) for more details.