Skip to main content
. 2017 Jan 15;145(Pt B):337–345. doi: 10.1016/j.neuroimage.2015.12.050

Fig. 2.

Fig. 2

Multiple kernel learning (MKL) results. (A) Scatter plot between the real and predicted NA scores for the model based on patterns of brain activation to threat stimuli in the directed away context. The correlation and the MSE between the real and predicted NA scores were r = 0.52 (p-value = 0.01) and MSE = 24.43 (p-value = 0.01). (B) Weights per voxel in a whole brain fashion, the color bar represents the full range of the weights. The images show the coronal, sagittal and axial slices (MNI coordinates). (C) Top 6 regions ranked by the MKL model that were relevant to the prediction. The percentage of weight of each region are in parentheses. The colors represent the weights per voxel within each region, including the right middle occipital gyrus, right middle temporal gyrus, right caudate nucleus, right inferior frontal gyrus (orbital part), right medial frontal gyrus and left insula. (2-column fitting image).