An interesting feature of human beings is that they are capable of continuous learning: they gather, learn from and build on experiences and acquired knowledge throughout their life spans. Machine learning researchers are trying to design systems that can do much the same. These systems would be able to perform new tasks using knowledge from those they’ve seen or performed before.

However, designing these new intelligent systems is not without its challenges—and as Canada Research Chair in Lifelong Machine Learning, Dr. Sarath Chandar Anbil Parthipan is trying to come up with new architectures and algorithms that could overcome some of these. He and his research team are also translating advances in lifelong learning to develop dialogue systems, drug discovery algorithms, and medical artificial intelligence systems that can accumulate knowledge over time and adapt as the world changes.