Let us now go back to our example in section 1.2.3 where we estimated the price of a used car. There we used a single input linear model (2.15) g(x) = wix + Wo where wi and wo are the parameters to learn from data. The wi and wo values should minimize N (2.16) E(WI, WolX ) = E[rt - (wix + wo)]2 2 1 = 1 Its minimum point can be calculated by taking the partial derivatives of E with respect to wj and wo, setting them equal to 0, and solving for the two unknowns: Extrt - XIN W1 Er(x1 ) 2 - NX- (2.17) WO = F - WIX