Gardner and Derrida have introduced a natural version of the problem of the capacity of the binary perceptron "with temperature," and they proposed (based on "physical" methods) remarkable formulas ...
Results that may be inaccessible to you are currently showing.
Hide inaccessible results
Feedback