Statistical tests for classification problems

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

Statistical tests for classification problems

asadbtk
Hello Eibe,

For instance, we have the regression problem in weka and our actual minus predicted values (absolute residuals) for algorithm A is 0.3,0.1,0.02,0.5 etc and same as some values for algorithm B. Based on that data we can perform statistical test (not t test in weka, but other tests in excel or other editors). We can do statistical tests easily by calculating its mean, median etc. However, if we have similar problem for classification, we gets the residuals or errors like 0,1,0,01, etc for both type of algorithms, then how can we perform such tests on classification problems? 

Thanks 

_______________________________________________
Wekalist mailing list -- [hidden email]
Send posts to: To unsubscribe send an email to [hidden email]
To subscribe, unsubscribe, etc., visit
https://list.waikato.ac.nz/postorius/lists/wekalist.list.waikato.ac.nz
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html