Figure 2.
A. An illustration describing the construction of the predictor matrix. Each row contains the time course data and contextual information from one voxel, while the ground truth is a binary array equal to 1 if the voxel is expected to be active for a given layer assignment and stimulation paradigm or equal to 0 otherwise. B. Flow diagram outlining how the eight animals were used for training, cross-validation, and statistical testing.