1.1.3.  \(f_1\ =\ C_d\left(0_d,\ I_d\right)\ and\ f_2\ =\ C_d\left(\left(0_{d-1},\ d\right),\ I_d\right)\)

\(\pi_1\ =\ 0.5\ \&\ \pi_2\ =\ 0.5\)

(B) LDA and QDA

For LDA, the training error decreases as dimension increases but the Test error almost remains same.