Employing ML in security in the real world
The real world is ever-changing and quite messy. You may think that there is a straightforward simple solution to a problem, but it’s unlikely that the solution to any given security problem is either straightforward or simple. What you often end up with is a layering of solutions that match the requirements of your environment. Consequently, you might find that an ML application designed to detect threats is part of a solution, the flexible part that learns and makes a successful attack less likely. However, you likely need to rely on traditional security and service-based security as well. It’s also important to keep user training in mind and not neglect those small things.
The reality of ML is that it’s a tool like any other tool and not somehow a magic wand that will remove all of your security problems. If Chapter 3 shows you anything, it demonstrates that ML security exploits exist in great quantities and that...
