Consider the data in Table $2.$

Object S C F V

o$1$ M B b Y

o$2$ S R s Y

o$3$ L G p Y

o$4$ L G s Y

o$5$ S R w N

o$6$ L R w N

o$7$ L R p N

Table $2$: Objects with attributes and their classification as V $($Y$,$ N$).$

We have data for another object, as shown in Table $3,$ with an unknown classification.

Object S C F V

o$8$ L G w $?$

Table $3$: Object o$8$ with unknown classification.

$($a$)$ Use the Hamming distance measure $($see Appendix on p$.7)$ to calculate

the distance between o$8$ and the $7$ instances in Table $2.$ Show the resulting

ordering. $(5)$

$($b$)$ If we want to use kNN to classify the new instance, o$8,$ what would be a good

value for k$?$ Justify your answer. $(4)$

$($c$)$ Use kNN with k $=3$ to classify the new instance, o$8.(4)$

$($d$)$ Discuss what effect choosing k $=4$ would have on the classification of the

new instance, o$8?(2)$

F$15$