Rapid Miner Session

1
Rapid Miner Session

• Three steps for use
• Assign the dataset file first
• Select functionality
• Execute function using Rapid Miner
• Sample Datasets
• You can use the ARFF format datasets provided in
  course website.
• Or, Proben Datasets from UCI, you need to change
  the data format to ARFF before use.
• Sample dataset link http//www.cis.syr.edu/hhuan
  g13/cis600/notes/RM.ppt
• Course website link http//www.cis.syr.edu/hhuan
  g13/cis600/index.html

2
Step 1

• Download and Install Rapid Miner
• http//rapid-i.com/content/view/130/82/
• Create a Rapid Miner Project ( click new icon )

CIS 600 Analytical Data Mining,EECS, SU
3
Step 1
Setting data source

• Set a data source for project , usually it is a
  data file( arff format or excel spread format),
  we use iris data file as an example.
• In the left side, right click on the Root icon,
  and choose the ArffExampleSource menu to select
  input datasets as below.

CIS 600 Analytical Data Mining,EECS, SU
4
Step1
Setting data source(cont)

• Assign the data file to data source

CIS 600 Analytical Data Mining,EECS, SU
5
Step1
Setting lable_attribute

• Assign label_attribute to arffExampleSource, this
  parameter will let RapidMiner know which
  attribute in your data source is the class
  attribute.

CIS 600 Analytical Data Mining,EECS, SU
6
Step2
Select functionality i.e analysis type

• Choose one of the data mining analysis type for
  project, such as clustering, classification,
  decision tree, etc.
• For example, we create a choose New Operator-gt
  Learner -gtSupervised -gtWake-gtTress-gtW-J48 to
  create a W-J48 operator. This is decision tree
  functionality.

CIS 600 Analytical Data Mining,EECS, SU
7
Step2
Decide Parameters

• Decide the detail parameters for clustering.

CIS 600 Analytical Data Mining,EECS, SU
8
Step3
Execute

• Execute . click the button in top side to run
  the clustering.

CIS 600 Analytical Data Mining,EECS, SU
9
Step3
Check results text view

• Check results from output panel. There are
  different output panels if you choose different
  analysis function.

CIS 600 Analytical Data Mining,EECS, SU
10
Step3
Check results graph view

• Graph view

CIS 600 Analytical Data Mining,EECS, SU