There are 4 important terms in Confusion Matrix,you can easily calculate them with due attention to their concept.

False Positives : The cases in which we predicted YES and the actual output was NO.

False Negatives : The cases in which we predicted NO and the actual output was YES.

True Positives : The cases in which we predicted YES and the actual output was also YES.

True Negatives : The cases in which we predicted NO and the actual output was NO.

I hope my answer can be useful.

Narges Takhtkeshha

Yes, Your answer is useful.

It is simple if we have and if we work with a binary classifier. But, what if we work with a multiclass classifier?

How can we determine values for TP, FP, FN and TN in that case?

Multiclass classifier could be one 4x4 or 9x9, for example.

Kind regards,

Zeljko Vujovic

Hi

It can be considered as a tow class confusion matrix. For example suppose that there is 5 classes(a,b,c,d,e) . TP, FP, FN and TN for class a and b can be computed in this way:

for class a : class1 = a; class2 = b, c, d, e

for class b: class1 = b; class2 = a,c,d,e

and so on.

Hope this is useful ;)

Hi, Alireza

That seems interesting, unusual but also incomprehensible

What is a tow class confusion matrix?

How it is defined?

Kind regards,

Hi Zeljko

https://www.dataschool.io/simple-guide-to-confusion-matrix-terminology/

I hope this link works.

best regards

Hi Alireza,

I reviewed that link earlier. It does not answer the question: what is towing a matrix of confusion?

Idea is incomprehensible:

for class a: class 1 = a; class 2 = b, c, d, e

for class b: class 1 = b; class 2 = a, c, d, e

have a 9x9 matrix.( it is in Attachment)

There are 9 real classes in the vertical columns, and there are 9 predicted classes in the horizontal rows. I am attaching a spreadsheet that shows this.

Therefore, the explanation you provided is unclear.

There is also no answer to the question: what is a tow confusion matrix?

Kind regards,

Željko Vujović

Confusion matrix 9x9 is on the link below, also: https://stats.stackexchange.com/questions/179835/how-to-build-a-confusion-matrix-for-a-multiclass-classifier

In a 9*9 confusion matrix you can use some measures to get the accuracy of classification like producer's and user's accuracy and F-score that are as the same as TP,FN, ... in the content. You can also imagine 9*9 matrix to be a 2*2 (class 1 against other classes or class 2 against other classes and so on)