At Virginia Commonwealth University (VCU), Milos Manic, Ph.D., and colleagues develop machine learning (ML) systems that have the ability to solve problems by learning on their own, directly from data instead of explicit programming.  ML research is one of the hottest fields in computer science in part because it offers the capability of computer systems to proactively respond in a useful and intelligent manner.  Familiar consumer-facing applications of ML-based systems are the "virtual assistant" systems of Amazon's Alexa or Apple's Siri, but ML-based systems have disrupted or are soon to disrupt every industry.  Manic and colleagues have recently received recognition for their application of ML to cybersecurity.

AICS – An “artificial intelligence breakthrough”
An Idaho National Laboratory (INL) researcher and inventor, Manic and colleagues Todd Vollmer and Craig Rieger were recently awarded the Federal Laboratory Consortium Award (FLCA) in the Far West Region.  This Outstanding Technology Development Award recognized the team’s work on the Autonomic Intelligent Cyber Sensor (AICS), described as an important artificial intelligence breakthrough. Developed to protect the nation’s critical infrastructure from devastating cyber attack, AICS can identify and divert hackers without human intervention. Working autonomously through ML-based techniques to identify and map industrial control systems, AICS recognizes anomalous network traffic indicative of a cyber attack, alerts human operators to the problem, and can frustrate hackers by deploying virtual decoys.

The concept of machines learning without human intervention can bring apprehension for some, stemming from the widely-debated public debates associated with artificial general intelligence. Manic observes that, as our economy and lives become increasingly dependent on software and the connectivity of the Internet of Things (IoT), ML can help manage risk by becoming part of the immune system that monitors and protects our world. Manic asks, "How do we trust these machines that they will be able to do the right thing?” These machines are “as smart as you enable them to be. The algorithms we create these days are capable of making decisions on their own. We are not hard coding their behavior anymore.”  ML-based systems like AICS will play a key role in protecting the most critical government infrastructure systems that are in use daily, such are nuclear power plants.  These systems demand the most sophisticated cyber-physical protection.  Recognizing the current and future needs to protect our nation’s infrastructure, the Commonwealth of Virginia includes Cyber and Cyber-Physical Security as a Strategic Growth Opportunity Area for economic development and job creation.  AICS and other research projects in Manic's group are but one way that VCU is supporting the strategic needs of the Commonwealth.

Can you teach machines emotions?
Conversations about developing artificial general intelligence (AGI) increasingly raise questions about "emotional efficiency".  Humans are inherently social creatures.  Designing for our social nature, by providing AGI with emotional intelligence, can produce AGI systems that are more capable of recognizing and solving human problems.  Earlier this year in London, Manic spoke on his vision for a possible benevolent future for mankind's relationship with AGI at the 2018 IEEE Joint International Conference on Information Management and Processing and e-Society, e-Learning and e-Technologies (IEEE ICMP/ICSLT 2018).  Asks Manic, “how do you teach a computer to feel, love, dream, or forget?“

Manic welcomes the challenges of his research at VCU in machine learning, cybersecurity, energy security, infrastructure creation, fuel efficiency, virtual reality, and robotics. "I am really proud to be a part of this country and at VCU, we focus on the most critical challenges that our nation is facing today,” says Manic. He and his team relish the opportunity to solve hard problems and that will help bring a safer future.

About Professor Manic
Professor Milos Manic, Ph.D. is on the faculty of the Department of Computer Science and is the Director of the Modern Heuristics Research Group at VCU.  The Modern Heuristics Research Group focuses on machine learning, cybersecurity, and human-machine interfaces. Late last year, Manic opened VCU’s first Virtual Reality Laboratory (http://www.richmond.com/news/local/education/vcu-virtual-reality-lab-opens-to-awe-of-crowd/article_1c9e26e5-8749-527c-9be8-60d50ce1e8f9.html ) that can assist with interactive visual data mining , personnel training, education, and rehabilitation.  Manic joined Virginia Commonwealth University’s College of Engineering in 2014 and has since secured over $1.1MM in sponsored research for the Modern Heuristics Research Group. An incredible asset to the College of Engineering, Manic has published over 180 refereed articles, holds several U.S. patents, and is an expert in the subjects of data mining and data fusion, intelligent control, network security and infrastructure protection, robotics, and recognition. Manic is also leading a recently established VCU Cybersecurity Center comprised of programs in Computer Science, Electrical and Computer Engineering, and Homeland Security and Emergency Preparedness.