Consider the following dataset of 10 training examples:
ID CPI A1 A2 A3
1 1.52 27.82 0.40 59.61
2 3.00 44.49 10.10 75.77
3 3.77 54.69 7.20 73.12
4 8.67 32.56 14.20 80.99
5 2.86 30.77 5.30 70.48
6 8.05 28.31 12.00 80.24
7 1.80 59.21 3.40 45.00
8 7.54 34.28 11.50 80.15
9 5.81 39.20 12.50 81.30
10 2.23 50.10 6.70 53.68
CPI A1 A2 A3
(a) Discretize the continuous attributes ($A_1$, $A_2$, $A_3$, $CPI$) to categorical using
equal-depth (frequency) binning, two bins for each attribute (e.g., 1 and 2).
Write the results in the table to the right.
(b) If a Bayesian Belief Network showing the conditional dependence
relationship between the features as follows:
$A_1$
$A_2$ $A_3$
$CPI$
Given a new query ($A_1 = 40.81$, $A_2 = 13.10$, $A_3 = unknown$), what category of CPI
will it be classified? Show calculation details based on conditional probability
regarding the training data in (a)?
