2 ms·
Or you could think of each NB classifier as a convolutional filter, followed by maxout / selection by voting at the top.
by visarga 10y ago
Or you could think of each NB classifier as a convolutional filter, followed by maxout / selection by voting at the top.