CYBER ATTACKS EVALUATION USING CONFUSION MATRIX
This article will make you understand confusion matrix and it’s usage in cyber security.
Let’s first understand the term confusion matrix -
The term confusion matrix in machine learning is used in prediction analysis. To put it in simple word the main goal of a machine learning model is to give high accuracy in prediction. So to measure the performance of the prediction we use confusion matrix.
The attribute that we get after the prediction of the confusion matrix are True Positive, False Positive, True Negative, False Negative which are used in visualizing important predictive analytics like recall, specificity, accuracy, and precision.
Now consider a scenario in which a company build a anti-virus which uses Artificial intelligence to detect a malicious file/document containing malware or a virus and you are using it in your office PC, while using your office PC you get malicious file/document form anonymous person containing virus, considering that the al anti-virus detect no virus through that file/document and gave you a “FALSE POSITIVE”.
But If you get a resume and the anti-virus detect a virus giving you a “FALSE NEGATIVE ”.
And the Fact that AI based anti-virus detect the virus when you receive a virus containing document or you are trying to open it which means it’s gives you “TRUE POSITIVE” and also if the anti-virus detect no virus inside the document which it doesn’t have in the first place than it gives “TRUE NEGATIVE”.
“FALSE POSITIVE” and “FALSE NEGATIVE ” count predicted by the AI based anti-virus are know as “TYPE 1” and “TYPE 2” errors.
Type 1 error : Type 1 error is the dangerous error where you get the result positive even though something is going wrong in your PC. This error gives the hacker the privilege to access the PC which is security breach.
Type 2 error : Type 2 error is less dangerous than Type 1, where you get the result negative even though the file doesn’t contain any virus which let to further investigation of the person after getting the foolproof of their innocence.
That’s it for now. Thank you everyone for reading the article.