Hello,
I m having difficulties finding the right probability distribution for the matrix "items".
The way I see the problem is that I need to place f*N active neurons in a set of N neurons, such that for each pattern, I always have the same fraction f of active neurons. But when using a probability distribution like we are expected to, the number of active neurons is not exactly f*N and is oscillating. Are we expected to have exactly f*N neurons for each of the L patterns?