nave bayes

barkhordari55

hello 

please help me

I don't understad the value in the distribution table in nave bayes?

 

lionelderkrikor

Hi @barkhordari55,

 

Here, you can find an Excel file using the "Golf dataset" to better understand how the Naive Bayes algorithm (without Laplace correction) works : 

 

https://drive.google.com/open?id=12mELZ_SW8fv-VfeRkY-mUjqEUb42ODx6

 

I hope it helps,

 

Regards,

 

Lionel

barkhordari55

Hello Lionel, thank you very much
It was very useful
I have another question
Can you interpret the following results from my RapidMiner?

barkhordari55

barkhordari55

Hello
please interpret the following results from my nave bayes in RapidMiner?

lionelderkrikor

Hi @barkhordari55,

 

Difficult to interpret raw results from these screenshots. (I don't know what are your attributes and your target variable)

Maybe I would better understand if you share your dataset(s)...

 

Regards,

 

Lionel

lionelderkrikor

Hi again @barkhordari55,

 

I would say (but to be confirmed by your dataset(s)) that you have an imbalanced dataset : You have a (very) big majority of

[target = True] and a minority of [target = False]. So a priori the model you builded consider the prediction  = [target = True] whatever

the value of your attributes. It is often the case when a dataset is imbalanced. In fine you have a (relativ) good accuracy but a very bad recall for [target = False] => You are not able to detect and predict the cases [Target = False]

If your primary goal is to detect and predict the case [Target = False], you have to pre-process your data, to increase the performance of your model.

 

Regards,

 

Lionel

 

NB : To be confirmed by your dataset(s)

barkhordari55

hello lion

thank you very much