I’ve been experimenting with the separation of two classes of objects in a 2D space. The data is treated as a bitmap image, allowing for convolution that melts away overfitness.
Does anyone have any suggestions on what this could be useful for?
