The applet above implements a neural network (visualised on the top right) that finds an appropriate weight distribution using a genetic algorithm (visualised by the bar chart), enabling it to emulate one of the boolean functions AND, OR, XOR and IF.. THEN... You can change the parameters of the genetic algorithm by specifying the new value in the text boxes and subsequently pressing the submit button. You can change the boolean function the network will emulate by pressing the buttons labeled AND, OR, XOR, IF.. THEN... Different parameter values will change the behaviour of the algorithm. Try to optimise them for a quick convergence to a valid solution without risking to get stuck in local optima. Sorry for the cheesy interface, it's one of my earlier projects...