Machines That Learn Like Humans
The rapid advance in artificial intelligence has been equated to the Industrial Revolution and the development of the Internet: that is, it has the potential to impact every sector, with far-reaching social and economic implications. At the core of these advances is machine learning—a field between science and engineering focused on algorithms that allow machines to learn. But despite widespread investment, both the “hardware” and “software” used by intelligent machines still lags behind that of humans.
Dr. Graham Taylor, Canada Research Chair in Machine Learning, wants machines to learn more like humans. He envisions a next generation of systems that will learn from very few examples and be able to explain their decisions. Taylor wants to remove the human factor from prediction by creating algorithms that are more systematic, accurate and affordable.
Taylor also plans to widen the scope of these systems to help more people on platforms ranging from servers to phones. He and his research team are focusing on applications that enable computers to “see”—from distinguishing among insect species to understanding human and animal behaviour.
Taylor and his team are developing open-source software and outreach strategies to support the use of machine learning across disciplines. By developing better systems and bringing this technology to reality, Taylor’s research will help realize machine learning’s full potential.