3. The logistic function is defined as a ) 1 A) y(x)= 1+e-* B) y(x)= 1 1+ e* 1 C) y(x)=; 1-et 1 D) y(x)=1 x= 1-e 9 4. Given the input vector X" = (x,x2,....,) and the linear model Y=X"B, minimize the residual sum of square RSS(B) =(Y - X"B)?, the best value of Bis ) A) B = (X" x) 'X'Y B) B= X'Y C) B =(x)'X'Y D) B =(X")"XY 5. The following table shows both of the training error and the testing error of four regression models. Which model is the best? ( ) Model A Model B Model C Model D Training error 0.32 0.15 0.14 0.10 Testing error 0.35 0.17 0.28 0.64