Tài liệu Wiley - Data Mining with Microsoft SQL Server 2008 (2009)02 pptx

Maclennan c01.tex V2 - 10/04/2008 1:59am Page 3

Introduction to Data Mining in SQL Server 2008 3

Figure 1-1 Student table

In contrast, the data mining approach for this problem is almost the reverse

of the query-and-explore method. Instead of guessing a hypothesis and trying

it out in different ways, you ask the question in terms of the data that can

support many hypotheses, and allow your data mining system to explore them

for you.

In this case, you indicate that the columns IQ, Gender, ParentIncome,

and ParentEncouragement are to be used as hypotheses in determining

CollegePlans. As the data mining system passes over the data, it analyzes the

influence of each input column on the target column.

Figure 1-2 shows the hypothetical result of a decision tree algorithm operat￾ing on this data set. In this case, each path from the root node to the leaf node

forms a rule about the data. Looking at this tree, you see that students with IQs

greater than 100 and who are encouraged by their parents are highly likely to

attend college. In this case, you have extracted knowledge from the data.

As shown here, data mining applies algorithms such as decision trees,

clustering, association, time series, and so on to a data set, and then analyzes

its contents. This analysis produces patterns, which can be explored for

valuable information. Depending on the underlying algorithm, these patterns

can be in the form of trees, rules, clusters, or simply a set of mathematical

formulas. The information found in the patterns can be used for reporting (to