Combining the unique strengths of humans and machines for better security outcomes
Today’s security landscape is changing very fast. The number of cyberattacks each day has risen from a mere 500 to an estimated 200,000-500,000. The volume of threats and information that must be processed is greater than humans alone can manage. We need the speed of machines to process, adapt, and scale.
But we need humans too, to match and outmatch the wits and ingenuity of the human attackers on the other side of that code. In short, we need teams of humans and machines, learning and informing each other—and working as one.
McAfee has fully embraced security analytic solutions using advanced, adaptive, and state-of-the-art machine learning, deep learning, and artificial intelligence techniques. Driving the pace of innovation, McAfee is moving quickly to evolve beyond the standard forms of advanced analytics to adopt a multi-layered approach known as “human-machine teaming.” This approach, by adding the human-in-the-loop within our products and processes, shows a 10x increase at catching threats with a 5-fold decrease in False Positives.*
* MIT 2016, Kalyan Veeramachaneni and Ignacio Arnaldo, “AI²: Training a big data machine to defend”.
Layers of artificial intelligence for each level of task
As humans have defined and refined advanced analytics, data scientists and technologists have recognized that there is an evolution of complexity towards more predictive and cognitive forms of computing. These levels, as depicted here, build one upon another toward the goal of better and faster intelligence.
Evolving machine learning for a better threat defense
Recent research highlights the need for machine learning for advanced detection capabilities. McAfee is evolving its machine learning cybersecurity technology to even more complex analytics called deep learning and artificial intelligence. Deep learning is the machine learning-based analytics approach that uses many layers of mathematical neurons—much like the human brain. It provides reasoning and a feed-forward or backward convolution of decision-making. Artificial intelligence adds complexity to deep learning, appending reasoning, suggested actions, and problem solving, often working in an n-dimensional space (like the brain). Machine learning, deep learning, and artificial intelligence are mathematically more complex as the computation becomes more brain- and human-like.
Putting the human-machine team to work
Machine learning and deep learning are pervasive within the McAfee product portfolio. McAfee Endpoint Security suites feature the machine learning capabilities of Real Protect. Our suite of advanced threat analysis solutions, McAfee Advanced Threat Defense and McAfee Cloud Threat Detection, use deep neural networks to provide advanced malware behavior analysis. Learn more about our offerings featuring machine learning and deep learning capabilities below.
Learn more about machine learning and artificial intelligence
This white paper draws an analogy between machine learning and analysis of the behaviors and characteristics of autonomous cars in various scenarios.