Confusion Matrix

#1
Hi everyone,

I have a question involving confusion matrix's I'm unsure on how to answer.

A programmer runs 1000 files of his Java code through a code analysing tool. This tool predicts that 37 of these files contain a bad code smell. It is discovered that

a. Only 25 of the 35 files predicted actually contain are defective
b. An additional 12 of the 1000 files (not predicted by the tool) are subsequently found to actually to be defective.

Construct a Confusion Matrix showing this performance of the tool.

The confusion matrix will look something like this :


Predicted defective Predicted defect free
Observed defective
Observed defect free

Any help at all would be appreciated.

Thanks
 
#2
Hi everyone,

I have a question involving confusion matrix's I'm unsure on how to answer.

A programmer runs 1000 files of his Java code through a code analysing tool. This tool predicts that 37 of these files contain a bad code smell. It is discovered that

a. Only 25 of the 35 files predicted actually contain are defective
b. An additional 12 of the 1000 files (not predicted by the tool) are subsequently found to actually to be defective.

Construct a Confusion Matrix showing this performance of the tool.

The confusion matrix will look something like this :


Predicted defective Predicted defect free
Observed defective
Observed defect free

Any help at all would be appreciated.

Thanks
It is a simple 2x2 table! The observed defectives is nothing but actual no. of defectives which can be obtained using (a) and (b).
Predicted defective is given in your problem statement.
Just carefully look at the definition of confusion matrix http://en.wikipedia.org/wiki/Confusion_matrix