Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

Can anyone explain what the Bayes error rate is in Naive Bayes, for instance in matlab:

[class, err]  = classify(UnseenTestdata, training_data, target_class,'diaglinear')

I get an error rate (err) of 0.5040 what does this mean? Im not a math orientated person so plain english would help. Is this good, bad or very ugly?

share|improve this question

1 Answer 1

From the matlab help function:

Err is an estimate of the misclassification error rate that is based on the training data. CLASSIFY returns the apparent error rate, i.e., the percentage of observations in the TRAINING that are misclassified, weighted by the prior probabilities for the groups.

So, if 0.5 is close to your guessing rate you are doing poorly, otherwise it's not bad.

share|improve this answer
    
How can you say your doing poorly but then say otherwise its not bad? At 50% accuracy across all classes its kinda rubbish. I had been thrown off because my accuracy from the confusion matrix was 81.49% but I had noticed that it classified the biggest body from the testdata correct (one class) while leaving alot of the other classes with almost 0% accuracy. –  Garrith Graham Nov 19 '12 at 16:28
    
Not sure if i understand correctly, but as I read it you have the output of two methods. In this case you can just compare the results manually and determine which one is better. –  Dennis Jaheruddin Nov 19 '12 at 16:41

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.