Confusion Matrix & It’s Role In Cyber Crime Cases

In this world of Artificial Intelligence which is created through Machine learning and Deep learning, making predictions(earlier which was only in the hands of humans) is the power that machines get, accuracy of the model matters a lot, but just achieving the high goals of accuracy is enough?

Well no, in real life there are certain situations between the hypothesis and the null hypothesis.

Hypothesis is which is “supposed” and null hypothesis is which is “reciprocal of that supposition”.

Confusion matrix is the one which clearly signifies these situations.

This matrix works for the classification type of dataset, it has two rows and two columns as you can see.

This matrix represents the occurrence of an event which will exist among these four conditions.

Lets take an example of a covid test reports.

Here hypothesis is “person is infected” means reports are positive, and null hypothesis is “person in not infected” means negative.

So, if two persons are diagnosed by AI machine (and let person A is infected in real), if reports are positive then it is “True Positive”, if reports are negative then it is “False Negative”

(Let person B is not infected) and his reports are positive then it is called “False Positive” and if reports are negative then it is called “True Negative”.

These “False Positive” and “False Negative” are the type 1 and type 2 error respectively.

In cyber crimes many times these errors are probable to happen something which is not desired has happened due to the these type of error like some message send to the unintended system(which is predicted as true by machine), hence False Positive error generate and at times come ,when an intended message is not send to the desired system( which maybe predicted as null hypothesis by the machine) then, it is called False Negative error.

Thanks for reading :)