Hi all,
I am interested to classify a dataset with the MultiLayerPerceptron in WEKA experimenter (using 10-folds cross-validation; 10 iterations). I have 10 classes and I was wondering how to obtain (or calculating) from the output file the confusion matrix showing the CCI for each group (as I can obtain in the output of the explorer).
Moreover, I was wondering if you might give me more details about the ROC, k-stat and TPR and FPR scores. Do not these measures refer to a binary classification? How are these calculated with a 10 classes output?
Finally, can I rank variables (corresponding to the input nodes) according to their importance or "weight" for classification?
Thank you in advance for your help!!!!
Cheers
I am interested to classify a dataset with the MultiLayerPerceptron in WEKA experimenter (using 10-folds cross-validation; 10 iterations). I have 10 classes and I was wondering how to obtain (or calculating) from the output file the confusion matrix showing the CCI for each group (as I can obtain in the output of the explorer).
Moreover, I was wondering if you might give me more details about the ROC, k-stat and TPR and FPR scores. Do not these measures refer to a binary classification? How are these calculated with a 10 classes output?
Finally, can I rank variables (corresponding to the input nodes) according to their importance or "weight" for classification?
Thank you in advance for your help!!!!
Cheers