This chapter has illustrated the different techniques that can be used to increase the effectiveness of user authentication procedures and to promptly detect the possible presence of compromised user accounts.

These techniques are based on the use of biometric evidence, such as facial recognition, or biometric behavior, such as keyboard dynamics, which can be implemented using AI algorithms that exploit neural networks, such as MLP.

We have also seen how to reduce the dimensionality of the data to their main components by using PCA.

Finally, the advantages and disadvantages associated with the use of biometric evidence for user authentication and recognition were highlighted.

In the next chapter, we will learn about fraud prevention using cloud AI solutions.