Hands-On Artificial Intelligence for Cybersecurity

HMM example

In our example, the possible observations are as follows:

ob_types = (‘W’,‘N’ )

Here, W stands for Working and N for Not Working, while the hidden states are as follows:

states = (‘L’, ’M')

Here, M corresponds to Malicious and L corresponds to Legitimate.

The sequence of observations comes next, which is associated to the single instructions that get executed by the program:

observations = (‘W’,‘W’,‘W’,‘N’)

This sequence of observations tells us that after the execution of the first three instructions of the program, the machine worked properly, while it stopped working only after executing the fourth instruction.

On the basis of this sequence of observable events, we must proceed with the training of the HMM. To this end, we ...