a.) P(C,R,S,W) = P(C) x P(R|C) x P(S|C) x P(W|S,R) = 0.5 x 0.8 x 0.1 x 0.99 = 0.0396
b.) P(¬C,¬R,S,W) = P(¬C) x P(¬R|¬C) x P(S|¬C) x P(W|S,¬R) = 0.5 x 0.8 x 0.5 x 0.9 = 0.18
, i.) Yes
ii.) Yes
iii.) No
iv.) No
X1 X2 X4 X5
Classification
, Q1.)
One advantage of using the KNN classifier is that it is a transparent model, easy to use. Classification
quickly computed and no model assumptions made.
One disadvantage of using the KNN classifier is changing the k parameter / distance metric /
weighting function can have a big effect on the classification results.
Q2.)
Accuracy = (TP + TN) / (TP + TN + FP + FN)
Classifier 1 = (12+5) / (12+15+8+5)
=
=0.425
Classifier 2 = (17+16) / (17+4+3+16)
=
=0.825
a.) Classifier 2 more accurate.
b.) Sensitivity: TP / (TP + FN)
Specificity: TN / (TN + FP)
b.) P(¬C,¬R,S,W) = P(¬C) x P(¬R|¬C) x P(S|¬C) x P(W|S,¬R) = 0.5 x 0.8 x 0.5 x 0.9 = 0.18
, i.) Yes
ii.) Yes
iii.) No
iv.) No
X1 X2 X4 X5
Classification
, Q1.)
One advantage of using the KNN classifier is that it is a transparent model, easy to use. Classification
quickly computed and no model assumptions made.
One disadvantage of using the KNN classifier is changing the k parameter / distance metric /
weighting function can have a big effect on the classification results.
Q2.)
Accuracy = (TP + TN) / (TP + TN + FP + FN)
Classifier 1 = (12+5) / (12+15+8+5)
=
=0.425
Classifier 2 = (17+16) / (17+4+3+16)
=
=0.825
a.) Classifier 2 more accurate.
b.) Sensitivity: TP / (TP + FN)
Specificity: TN / (TN + FP)
