When it comes to decision-making, VCU researchers are teaching computers to take humans out of the equation. Machine learning enables computers to make smarter, faster choices on the fly. But don’t get the wrong idea. This isn’t to take over the world; it’s to better work with it. Across campus, humans in computer science laboratories still provide the real brain power behind some of the industry’s most recent advancements in machine learning.
Milos Manic, Ph.D., professor in VCU College of Engineering’s Department of Computer Science, says humans and machines are already so interconnected that “everything is talking to everything.” His work in artificial intelligence (AI) is ultimately designed with people in mind, from bolstering national cybersecurity to improving the efficiency of a building’s energy. Manic is director of VCU’s Modern Heuristics Research Group, which uses computer intelligence to drive immersive experiences that help users virtually step into their own 3D data sets for a closer look at a problem.
Bartosz Krawczyk, Ph.D., assistant professor of computer science, heads VCU’s Machine Learning and Stream Mining lab. His research is focused on designing algorithms to mine and analyze high-speed data streams more accurately. He and his team are developing new machine learning methods that can autonomously improve their performance and adapt to changes and novel patterns in incoming data.
Alberto Cano, Ph.D., assistant professor of computer science, heads VCU’s High-Performance Data Mining lab. His research is focused on designing fast, efficient and accurate algorithms capable of scaling to big data and huge collections of information. He and his students have developed new machine learning methods for high-performance computing platforms that can extract knowledge from terabyte-size databases in academia and industry.
“In the upcoming decade. machine learning will blend seamlessly into our daily lives”Alberto Cano, Ph.D. Assistant Professor Department of Computer Science College of Engineering
Acknowledging that these technologies raise questions about how much to trust machines to do the right thing, these researchers emphasize that machines are as smart as you make them. Modern algorithms learn, act, improve themselves and are even capable of explaining their actions.