Home

10%

Lebo Radebe

MScAC Data Science Concentration alumna Lebo talks about her “one true love”: applying predictive models to solve real-world problems.

By U of T Statistical Sciences

Current Job: Data Scientist, The Hospital for Sick Children (SickKids)

Undergraduate Degree: Computer science and statistics (double major), African studies (minor), University of Toronto

Hometown: Pretoria, South Africa / Toronto, Canada

MScAC Cohort: 2017

LinkedIn Profile

⤙⟡⤚

Lebo took some detours on her way to data science, including a brief stint in economics, only to find her “one true love”: applying predictive models to solve real-world problems.

To pursue her passion, she enrolled in U of T’s Master of Applied Computing (MScAC) Data Science Concentration. Now a data scientist at SickKids, one of Canada’s major pediatric teaching hospitals, she is working on developing algorithms that help emergency doctors make the right decisions in how to best treat their patients.

“Our predictions inform doctor’s decisions and affect people's lives. It's not good enough to just say, ‘Well, that’s what the model told us.’ We need to understand the why behind the what, and how the process generating your data might have influenced your results.” - Lebo Radebe, Data Scientist, SickKids

Statistical Sciences: When did you first become interested in statistics and computer science?

Lebo Radebe: I took a bit of a detour before I landed on statistics and computer science. I started out in international relations and African studies, and then switched to economics. That’s when I took my first statistics course – and fell in love with statistics.

As for programming: I really got into that at the end of my fourth year when I took an internship at Citibank. My manager was very interested in natural language processing. I knew very little about that at the time, so I started reading textbooks and did online tutorials. That’s when I found out that data science and applying predictive models to real-world problems is my sweet spot.

After the internship, I targeted all my courses at the machine learning courses in computer science.

What made you decide on the MScAC Data Science Concentration?

The data science concentration was a perfect combination, allowing me to grow both my statistical knowledge, while working on my computer science skills.

Why was it important to you to focus on both – statistics and computer science?

The summer after I graduated, I worked with Anna Goldenberg at SickKids. That’s when I realized how important it is to combine computer science techniques with statistical best practices.

I was working on real data for the first time. It made me realize how much work goes into the statistics behind the models: cleaning your data and understanding data patterns. It’s easy to fall into a habit where you plug data into a machine, let it do some random number-crunching, and it spits out an answer. Our predictions inform doctor’s decisions and affect people's lives. It's not good enough to just say, “Well, that’s what the model told us.” We need to understand the why behind the what, and how the process generating your data might have influenced your results.

What did you work on during your internship at SickKids?

I worked – and am still working on – building a risk prediction dashboard that displays a patient’s risk for a number of things. One of the factors I’m looking at in particular is the risk of a patient returning to the emergency room within 72 hours.

The main goal is to help doctors make the right decisions for their patients and improve patient care, but also to decrease the burden on emergency doctors and cut down on waiting time.

What is the one thing you appreciated most about your internship?

The opportunity to build something for real-world implementation. When you take university courses, you try to understand the problem, maximize your accuracy, get an output, and make sure your model is working – and then you move on. It’s just so much more motivating when you know that your work will help make predictions for doctors in an emergency department. You take a lot more care when you know that the output has to predict well, because any false prediction will directly impact people.

What are the next steps for you?

I'm very happy with the work I’m doing at SickKids right now, but I’m still contemplating doing a PhD. Given that I have a brand-new baby, I probably wait a little longer to make that decision.

As far as long-term plans go: through the whole process of giving birth, I have become very interested in maternal health, specifically the role of midwives in reducing maternal and infant mortality rates. I’m originally from South Africa, and at some point, I would love to use my data science skills to strengthen the healthcare system back home, specifically in the area of giving birth. I always had it in the back of my mind that one day I might go back to South Africa to meaningfully impact people's lives by building solutions that can help people.

Interested in applying to the MScAC Data Science Concentration? Learn more about the program or connect with us at mscac@cs.toronto.edu if you have any questions.

We'd love to hear from you!

Button Label

Helen Li

Software Engineer at Microsoft Azure on how statistical thinking shows up in real-world systems.

Read

Professor Gwen Eadie receives COPSS Emerging Leader Award

What does leadership in statistics look like in statistics?

Read

Sissi Zhou

Actuarial science + statistics student Sissi talks about building confidence, making industry connections, and learning how actuarial work shows up in real business decisions

Read

Exploring the Universe: U of T Students’ Journey with the Galaxy Zoo Project

Undergraduate Statistics Students Contribute to Galaxy Zoo Project with Euclid Space Satellite

Read

Explore Scholarships and Opportunities in Actuarial Science

at the University of Toronto

Read

Kevin Haosui

Statistics PhD graduate Kevin talks about his career path, his postdoc at UC Berkeley, and Texas Hold’em.

Read