Supervised Maximum Likelihood Learning
		• Minimizing the squared residuals is equivalent to maximizing the log probability of the correct answer under a Gaussian centered at the model’s guess.
					d = the correct answer		y = model’s estimate of most probable value