Multiclass Sparse Bayesian Regression for fMRI-Based Prediction

Table 1

Simulated regression data. Explained variance 𝜁 for different methods (average of 15 different trials). The P-values are computed using a paired t-test.

Methods	Mean 𝜁	Std 𝜁	P-value to Gibbs-MCBR
SVR	0.11	0.1	.0 **
Elastic net	0.77	0.11	.0004 **
BRR	0.19	0.14	.0 **
ARD	0.79	0.06	.0 **
Gibbs-MCBR	0.89	0.04	—
VB-MCBR	0.04	0.05	.0 **

**Level of significance of the 𝑃-values between 0.01 and 0.05.