Practical Applications of Data Mining

162 Chapter 4 StatiStiCS for Data Mining

(Query 11)

Find attribute_1 times attribute_2 divided by attribute_3 from the table.

(SQL 11)

Select attribute_1 × attribute_2 / attribute_3 from the table.

These queries compute the value of the expected frequency based on

three groups of data: attribute_1, attribute_2, and attribute_3, where their size

must be the same. Finally, we define the c

2

-based query as follows, which

involves two subqueries, one for the observed variable and the other for the

expected variable:

(Query 12)

Are attribute_1 and attribute_2 independent in the table?

(SQL 12)

Select chi-square(∗) from the table where variable is observed and variable is

expected.

From the previous example dealing with price and feature attributes, w