Given the following FIT and TEST data sets:

FIT data set:

$=============$

Program Modules Actual no$.$ of Fault$-$Prone $($FP$)$ LOC

Faults $($Y$)$ Not Fault$-$Prone $($NFP$)$ Indep. Variable $($X$)$

$---------------------------------------------------------------$

A $0$ NFP $19$

B $1$ NFP $25$

C $0$ NFP $20$

D $2$ NFP $28$

E $1$ NFP $21$

F $3$ NFP $30$

G $2$ NFP $36$

H $5$ NFP $38$

I $7$ FP $41$

J $11$ FP $45$

K $16$ FP $50$

L $21$ FP $60$

TEST data set:

$=============$

M $??42$

N $??37$

Where $(?)$ means that we do NOT know its value and have to predict$/$estimate$.$

$($a$)$ Use CBR with Euclidean Distance as similarity function and un$-$weighted

average of THREE most similar cases to predict the number of faults in modules M

and N$.$ SHOW ALL YOUR WORK!

$($b$)$ Use CBR with Euclidean Distance as similarity function and Majority Voting

method with THREE most similar cases to classify modules M and N as FP or NFP

when C $=2\mathrm{.}25.$ SHOW ALL YOUR WORK!

$($c$)$ Repeat part $($b$)$ with Data Clustering method. SHOW ALL YOUR WORK!