Working of Naive Bayes Algorithm in Rapid Miner
Hello,
This might be easy but since I am new to RapidMiner any help will be appreciated. I am working with predicting results of a dataset using rapidMiner. I saw some youtube videos which showed x% validation that divides data in training set and apply model. I got that part. Rapid miner requires me to set label in the dataset ? What is this label? It also has restriction that label cannot be numeric. I am working with student data to predict success rate. How should I set label? How should I select attributes that maps to success rate. I am stuck at the point where I need to give proper input to Naive Bayes which gives me result that makes sense.I dont understand the criterion of performance vector.
Answers
-
Dear kijakarkar ,
your questions are quite diverse. Did you have a chance to have a look on our Getting Started page?
~Martin
0 -
Label is nothing but your target varible, you can specify which column you are trying to predict using the set role operator. Details are here http://docs.rapidminer.com/studio/operators/blending/attributes/names_and_roles/set_role.html
Also Naive Bayes can predict binary(yes/no) (true/false) or multi class predicitons, but not numerical
There are other algorithms that can predict numerical values.
As @mschmitz suggested, you will be extremely benefited from the gettting started videos, These are really good resources for getting started with Rapidminer and will help you understand how the utilize the platform for your specific cases
0 -
Thank you. I'll go through the getting started page0
Altair Employee