Database Reference
In-Depth Information
Table 7.4
contains a single record for a client who has a career in management, is
married, holds a secondary degree, has credit not in default, has a housing loan
but no personal loans, prefers to be contacted via cellular, and whose outcome
of the previous marketing campaign contact was a success. Is this client likely to
subscribe to the term deposit?
Table 7.4
Record of an Additional Client
Job Marital Education Default Housing Loan Contact Poutcome
management married secondary
no
yes
no
cellular Success
The conditional probabilities shown in
Table 7.5
can be calculated after building
the classifier with the training set.
Table 7.5
Compute Conditional Probabilities for the New Record
j
a
j
P
(
a
j
| subscribed =
yes)
P
(
a
j
( | subscribed =
no)
1 job = management
0.22
0.21
2 marital = married
0.53
0.61
3 education =
secondary
0.46
0.51
4 default = no
0.99
0.98
5 housing = yes
0.35
0.57
6 loan = no
0.90
0.85
7 contact = cellular
0.85
0.62
8 poutcome = success
0.15
0.01
Because
is proportional to the product of
times
, the naïve Bayes classifier assigns the class label
which results in the
greatest value over all . Thus,
is
computed for each
with
.
For
= {management, married, secondary, no, yes, no, cellular, success},
