From the course: Artificial Intelligence for Cybersecurity
Join today to access over 23,200 courses taught by industry experts.
Limitations and poor design
From the course: Artificial Intelligence for Cybersecurity
Limitations and poor design
- [Instructor] Machine learning just by its nature comes with limitations and dependencies. In addition, if implemented poorly it will lead your security teams to make a suboptimal or wrong decision when protecting from a threat. Machine learning is probabilistic. The algorithms, especially deep learning algorithms don't have domain knowledge. They don't understand the underlying network topologies the physics, or your business logic. These algorithms only look at input and output data and identify relationships between them without any meaning attached to those relationships. So it's quite possible that the train model that you created based on input data delivers results that violate the fundamental constraints of your environment. When implementing AI you want to bring subject matter expertise and deliberately add constraints to algorithms so they honor your rules and logic. Most of the machine learning algorithms lack…
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.