Skip to main content
Don’t miss our
best deal ever
offer ends july 23
save over $160
Sale ends in
6 months
Don’t miss our
best deal ever
for 6 months
save over $160
// //
Sponsor Content

Prof. Andrea Lodi is looking at utilizing mathematical optimization for extracting information from data.

There is a good chance that your smartphone is collecting data about you. It may track your movements throughout the day, count your steps, keep track of your networks, log your mobile shopping and banking activities, as well as your Angry Birds scores. More information about you may be coming from your fitness tracker, smartwatch, tablet, computer, TV and even the buildings you frequent.

But how is this information used? How can it be examined, selected and translated into tangible personal, organizational and societal benefits? These are questions Andrea Lodi is working to unlock. "Big data is certainly at the top of everyone's mind," he says. "But in order to see significant outcomes, it requires a certain degree of investment, both in people and resources."

For finding the right environment to advance what he calls "game-changers for the future of research," Prof. Lodi, Canada Excellence Research Chair (CERC) in data science for real-time decision-making – the biggest chair in Canada in the field of big data and operations research – relocated from Bologna, Italy, to Montreal last year.

Story continues below advertisement

Prof. Lodi, who has already weathered one winter in Montreal, says a number of favourable conditions prompted the move: the CERC investment, the calibre of his colleagues, and the commitment from Campus Montréal, which includes Polytechnique Montréal, HEC Montréal and Université de Montréal.

"In Montreal, we have this fantastic group that's made up of scientists working in applied math optimization and computer science, especially machine development dealing with data. And there is a strong connection to industry and organizations that are interested in following up," he says. "It's a unique opportunity that's not easy to find [elsewhere]."

While Prof. Lodi says he is still building his team, he has already seen "lots of good activity" towards his goals – which include advancing practical applications for his research.

A field where Prof. Lodi sees considerable potential for utilizing data to improve outcomes – both on the personal and organizational level – is health care. "We have a project that looks at many aspects, from scheduling operating rooms in hospitals to leveraging medical expertise for the best personalized patient care," he explains.

For coming up with the best strategy of care for an individual, Prof. Lodi suggests looking at essential historical data, such as dosages of drugs and strategies of care that have been applied to different patients, and then integrating the findings with personal data, such as age, past medical history, past reaction to medication, etc. "By interpreting the data that is already there, we can predict what types of strategies are going to be effective," he says. "We start out with a large number of strategies and then apply mathematical optimization to explore the best solutions."

What's new in this model is the sheer amount of real-time information, says Prof. Lodi. An IBM estimate predicts that online data exchange should surpass a zettaoctet – or one billion times the annual capacity of a domestic hard disk – in 2016. He adds that the data arrives from many different sources and platforms. For the medical application, for example, even a person's fitness app could yield data that her or his physician might find useful. "These things were completely out of reach in the past. Now we have the data but are not necessarily able to use it fully yet," he says.

Prof. Lodi and his team are developing models and algorithms that can help to rapidly access – and utilize – this wealth of strategic information. "Machine learning is about extracting information and catching patterns in the data that humans don't necessarily see because of the complexity. Once you have access to the knowledge, mathematical optimization can factor into the decision-making," he says.

Story continues below advertisement

In transportation, for example, big data can be useful for planning traffic flow, such as reacting in real time to traffic congestions or accidents, or helping commuters better co-ordinate their timing and routes through the city, says Prof. Lodi. Another application he and his team are working on provides valuable insights into customers' buying behaviours to the fashion industry for co-ordinating the types of collections available in stores.

The potential for putting big data to good use is vast, says Prof. Lodi, who adds that the outcomes of his team's work can affect every sector of the industry that uses optimization.

This content was produced by Randall Anthony Communications, in partnership with The Globe and Mail's advertising department. The Globe's editorial department was not involved in its creation.

Report an error
To view this site properly, enable cookies in your browser. Read our privacy policy to learn more.
How to enable cookies