CODEWITHSUNDEEP
javawekadecision-tree
Nipun Alahakoon
Nipun Alahakoon

Reputation: 2862

what is the meaning of the results saved in fast vector from Evaluation.predictions() Weka-api java

When i run the j-48 decision tree classifier in my data set, i evaluate it with fold cross validation and took the result into a fast vector by

predictions.appendElements(validation.predictions());

and i get the result as

NOM: 0.0 0.0 1.0 0.8349514563106796 0.1650485436893204
NOM: 1.0 0.0 1.0 0.8333333333333334 0.16666666666666666
NOM: 1.0 0.0 1.0 0.8333333333333334 0.16666666666666666
.....
......

what is the meaning of each column ?

Upvotes: 1

Views: 211

Answers (1)

Matthew Spencer
Matthew Spencer

Reputation: 2295

I'm not 100% sure about this, but the data that is displayed from the NumericPrediction class contains the following:

  • Actual Value
  • Predicted Value
  • Error
  • Weight
  • Prediction Intervals

It might be a good idea to look over the Validation Set you are using and then correlate any of the above attributes with the data that is present/predicted in the Validation.

Upvotes: 1

Related Questions