Learning
		¡ Inductive learning - based on past examples ¡ Learn a function h() that approximates real function f(x) on examples x ¡ Balance complexity of hypothesis with fidelity to the examples l Minimize α E(h,D)  + (1-α) C(h)