Table 3. The confusion matrix showing the strain classification using the front four principal components of PCA of the normalized data. The accuracy of the classification is defined as (252 + 240) / (290 + 299), namely 83.5%.



 

True

Class

 

 Predicted class

 ICV-K1 (287)

 T306 (303)

 % correct

 ICV-K1

 252

 38 

 86.9

 T306

 59

 240

 80.3

 


Supported by UNESCO / MIRCEN network