Christopher Whidden

Background

About Christopher Whidden

I'm an Assistant Professor in Computer Science at Dalhousie University. I develop efficient algorithms and software for NP-hard problems and Big Data, particularly in computational biology and oceans research. My research lab blends computer science theory and optimization to develop efficient usable software for applications ranging from high-confidence inference of the evolutionary history of bacteria to the automatic detection and species classification of at-risk fish using imaging sonar and video. My goal is to develop and optimize new algorithms that can process the massive amounts of data being generated throughout the world with a focus on bioinformatics and the ocean economy. Machine learning and data science are now actively embraced to solve practical real-world problems and it is crucial that we use algorithms with a solid foundation, actionable results, and efficient performance tailored to both large datacenters and small embedded devices. One major arm of my research is developing new algorithms and software for understanding complex evolutionary processes. For example, by leveraging new insights and a combination of new computational techniques I reduced the time required to compute gene transfer models called “maximum agreement forests” by six orders of magnitude. This research enabled comparing thousands of trees of the DNA from hundreds of species in fractions of a second rather than years. The other major arm of my research is collaborative applied research with ocean scientists and industry partners. With funding from Mitacs and computing resources from DeepSense, we help translate technology research and development to practical, responsible uses such as monitoring and protecting at-risk species of fish. For example, with the Offshore Energy Research Association of Nova Scotia (OERA) we developed a new method “echofilter” to automatically remove noise from underwater sonar data caused by the extraordinarily strong tides in the Bay of Fundy. This machine learning model is in active use saving 50% of the time needed to study fish health and safety around new tidal turbines, thereby promoting the development of green tidal energy. As another example, we developed a machine learning model with Innovasea to detect and classify fish species from underwater video. This research was recently deployed for field testing by Innovasea in collaboration with Nova Scotia Power at the White Rock Dam on Nova Scotia’s Gaspereau river, as featured by CBC News, Saltwire and Entrevestor.