M North Data Mining for the Masses 2012

Ενδεικτική Βιβλιογραφία • M. North, Data Mining for the Masses, 2012, ISBN: 978 -0615684376 • This book is licensed under a Creative Commons Attribution 3. 0 License

CRISP-DM Conceptual Model Data Mining for the Masses 3

https: //sites. google. com/site/dataminingforthemasses/ Elapsed_Time, Time_in_Community, Gender, Working, Age, Family, Hobbies, Social_Club, Political, Professional, Religious, Support_Group 8. 71, Short, M, No, 53, 1, 0, 0, 0 5. 24, Medium, F, No, 31, 0, 0, 0, 1, 1 4. 22, Medium, M, No, 42, 1, 1, 0, 0 4. 81, Long, F, No, 30, 0, 0 3. 95, Long, M, Yes, 29, 0, 0, 0, 1, 1, 0, 1 9. 35, Long, F, No, 40, 0, 0, 1, 0, 0 2. 91, Medium, F, Yes, 33, 0, 0, 0, 1 4. 54, Medium, M, Yes, 27, 1, 1, 1, 0, 0, 1, 0 7

A new data mining project in Rapid. Miner The Rapid. Miner start screen 9

Import Data Set 10

Import Data Set – Steps 5 11

How columns are separated 12

Names of the attributes 13

Data types, role 14

Where to store 15

Data View 16

Meta Data View 17

Toggle between Design Perspective and Results Perspective Design Perspective 18

Design Perspective 19

Drag and Drop 20

Data preparation: Select Attribute 21

Add select attribute operator to your data mining stream 22

Parameters pane and Select Attributes 23

24

Selected attributes for inclusion 25

Data preparation: Numerical to Binominal 26

Drag the Numerical to Binominal operator into your stream. 27

Play 28

Meta Data View: data type transformation 29

(Design Perspective) Frequency Pattern Analysis: FP-Growth 30

FP-Growth 31

FP-Growth - your DM stream Both your exa port and your fre port are connected to res ports 32

Parameters pane 33

Play 34

Religious organizations might have some natural connections with Family and Hobby organizations Further investigation

Create Association Rules operator 36

37

Play 38

Parameters 39

Changing the confidence percent threshold 40

Four rules found with the 50% confidence threshold 41

Do existing linkages between types of community groups exist? Yes, they do. We have found that the community’s churches, family, and hobby organizations have some common members. It may be a bit surprising that the political and professional groups do not appear to be interconnected, but these groups may also be more specialized (e. g. a local chapter of the bar association) and thus may not have tremendous cross-organizational appeal or need. M. North, Data Mining for the Masses, 2012