Options
Naive Bayes - RapidMiner Operator Reference manual
Contributor I
I was reading
through the Naive Bayes Model in the RapidMiner Operator Reference (pdf) found here -http://rapidminer.com/documentation/
and had a question.
In the Naive Bayes model (page 653/990) they are calculating the maximum
calculated probability for each label value.
1. They calculate the Posterior probability of label=Yes = >9/14,
2. Value from distribution table when Outlook = sunny and label = yes (i.e.*0.223*)
I understand posterior probability 9/14, How did they calculate 0.223? How did they arrive at this value?
Please let me know.
Thanks,
Ram
through the Naive Bayes Model in the RapidMiner Operator Reference (pdf) found here -http://rapidminer.com/documentation/
and had a question.
In the Naive Bayes model (page 653/990) they are calculating the maximum
calculated probability for each label value.
1. They calculate the Posterior probability of label=Yes = >9/14,
2. Value from distribution table when Outlook = sunny and label = yes (i.e.*0.223*)
I understand posterior probability 9/14, How did they calculate 0.223? How did they arrive at this value?
Please let me know.
Thanks,
Ram
Tagged:
0
Answers
the value from the distribution table for the case Outlook = sunny and label = yes will be calculated as 2 / 9 because there are 2 examples that fit the constraints having 9 times the occurrence of label = yes. The values in the operators manual do slightly differ because they have been computed using the Laplace option. If you disable this default parameter you should get a distribution table containing the expected values.
Cheers,
Helge
How the laplance correction work in this case to derive 0.2333?