In addition to clustering algorithms, it is possible to use classification algorithms for the detection of malware threats. Of particular importance is the classification of the malware carried out by using decision trees.
We have already met decision trees in Chapter 3, Ham or Spam? Detecting Email Cybersecurity Threats with AI, when we discussed the problem of spam detection. Now, we will deal with the classification problems solved by decision trees in the context of detecting malware threats.
The distinctive feature of decision trees is that these algorithms achieve the goal of classifying data in certain classes by modeling the learning process based on a sequence of if-then-else decisions.
For this characteristic, decision trees represent a type of non-linear classifier, whose decision boundaries are not reducible to straight lines or hyperplanes...