Table 3.

Benchmarking results: mean training time and standard deviation in seconds for different regression models.

Dataset	Toolbox	Regression model
		Ridge	Kernel Ridge	SVR (linear)	SVR (RBF)
MEG single-subjects	MVPA-Light	0.0016 ± 0.00006	0.019 ± 0.0001	–	–
	LIBSVM	–	–	0.02 ± 0.001	0.0041 ± 0.0002
	MATLAB	0.0061 ± 0.0002	–	0.018 ± 0.037	0.023 ± 0.0005
	Scikit Learn	0.0069 ± 0.0003	0.023 ± 0.003	0.654 ± 0.0647	0.481 ± 0.02
	R	0.055 ± 0.0027	–	1.59 ± 0.094	0.43 ± 0.002
MEG super-subject	MVPA-Light	0.015 ± 0.001	7.38 ± 0.023	–	–
	LIBSVM	–	–	0.653 ± 0.038	0.121 ± 0.014
	MATLAB	0.186 ± 0.007	–	6.931 ± 0.237	9.9798 ± 0.239
	Scikit Learn	0.062 ± 0.005	14.51 ± 0.21	3.213 ± 0.394	31.61 ± 1.51
	R	0.547 ± 0.0079	–	465.08 ± 49.83	151.66 ± 26.76
fMRI	MVPA-Light	0.165 ± 0.0042	2.026 ± 0.256	–	–
	LIBSVM	–	–	4.334 ± 1.48	2.819 ± 0.0412
	MATLAB	OOM	–	4.545 ± 0.353	4.563 ± 0.284
	Scikit Learn	0.638 ± 0.022	0.476 ± 0.01	16.138 ± 3.64	9.999 ± 0.59
	R	7.503 ± 0.593	–	37.211 ± 2.056	41.037 ± 2.298

For each combination of dataset and model, the fastest model is marked in bold. OOM, out of memory error; (p), primal form; (d), dual form.