Russ Greiner — Amii Fellow, Canada CIFAR AI Chair and one of the founding members of Amii — is being recognized for his outstanding contribution to furthering computing science with a Lifetime Achievement Award in Computing Science from CS-Can | Info-Can.

The award, announced May 10th, recognizes individuals who have made "outstanding and sustained contributions to Canadian computing over the duration of their careers."

"It's amazing to watch this field grow. I was at the first US AI conference," Greiner says. "I was at conferences when machine learning was just a workshop before we became a full-fledged entity. And given how successful this field has become, that genie is not going back in the bottle."

Greiner received his MSc and then PhD from Stanford University, and contributed to the Expert Systems movement, which aimed to develop artificial intelligence that mimicked the decision-making of human experts. In the 1990s, Greiner was among the first to explore issues in machine learning, with his research influencing the foundations of the fledgling discipline, including work on dealing with (even exploiting) missing data and actively learning models.

Greiner describes himself as "a bit of an opportunist" when it comes to his research interests, focusing on interesting challenges and datasets. He says he finds himself most drawn to the frontiers of ML, where there's space to experiment with new approaches and methods.

"I think my main skill is just looking at things critically and saying, ‘why are you doing that?’ So I see myself in that mode of thinking about how things could be done and not being stuck in just doing something in the standard way," he says.

"It's exciting because you get to explore directions that others haven't thought about yet."

Applications to Metabolomics, then Medicine

That spirit saw Greiner turn his attention to the intersection of machine learning, biology and medicine around the turn of the century. He started at the University of Alberta in 1997 , and initially explored many areas, from web design and recommendation to game design, as well as foundational work related to reasoning. Around this time, Greiner began working on his most visible project: using his machine learning expertise to contribute to the Human Metabolome Database (HMDB). The HMDB is the most complete collection of information about small molecule metabolites found in the human body. It has served as a valuable resource in medical advancements in pharmaceuticals, cancer research and other areas.

This project turned out to be a bridge to the medical community. Over the last 20 years, essentially all of his work has been in finding ways to apply ML to various medical tasks – ie, learning prognostic models that can predict various individual outcomes. For example, (1) whether a specific woman with breast cancer will experience a relapse within 3 years, based on information about the subcell location of various specific proteins in her cells, and other medical information; (2) how long a particular patient will live with a specified new liver graft; (3) whether a patient with depression will respond to a specific treatment; (4) what brain region will be irreparably damaged within 24 hours of an ischemic stroke; and (5) the likelihood of heart failure within a year, based on a cardiac Magnetic Resonance image.

More recent work relates to forecasting the number of Covid19 cases in a region in a few weeks – producing results that are comparable to the State-of-the-Art. All of these results, and many others, are in collaboration with many medical experts.

These interactions are often a consequence of his many dozens of public presentations summarizing how ML can be applied to medical tasks, leading to over 70 cold calls, which often evolve into specific projects. (Greiner acknowledges he still has not learned how to say "no".)

This has led to hundreds of relevant publications across many areas – around 45 in cancer, over 30 in psychiatry, 40 in medical imaging, etc., as well as many deployed in well-used websites. In 2022, he was awarded the Precision Health Innovator Award.

Despite his impressive achievements, Greiner is much more inclined to look towards the future rather than the past. His most recent work has focused on survivability predictions: using machine learning to look for patterns in medical data to help predict times to death or relapse, or disease onset. His research explores the potential for ML to help predict the survival times of people suffering from cancer, Covid19, cardiac issues and other conditions. While there are several nice results already – both foundational and applications – there is still more work to be done; Greiner is hopeful the work will lead to advances that will lead to better patient outcomes.

“I can imagine my tombstone may include a little survival curve or something," he laughs. "I think this is an area where my work will make a difference."

ML in AB

In addition to his own research, Greiner has been an influential figure in guiding the future of machine learning. He was one of four AI researchers who pitched the Alberta government to fund a research centre on machine learning. This led to the founding of the Alberta Ingenuity Centre for Machine Learning — now known as Amii — in 2002. Greiner would serve as the centre's first scientific director and be instrumental in turning Alberta into the hub for artificial intelligence research that it remains today.

"We had the insights to surround ourselves with the right people. We had the tenacity to think big. We said, 'Let's ask superstars like Rich Sutton, Dale Schuurmans, Micheal Bowling, and later, several others. Sure, they might say no, but let's try.'

We were lucky they all said Yes. And it just took off."

Greiner becomes the third Amii researcher and founder to be recently honoured by CS-Can | Info-Can. He will be presented with the award on June 7 at an awards gala in Montreal. The Lifetime Achievement Award was previously awarded to Rob Holte in 2021 and Jonathan Schaeffer in 2020.