“ML competitions aren’t the real world but are pretty close”: meet MG Ferreira, mathematician, entrepreneur, and Zindian

“I learned about Zindi at an online NVIDIA conference I attended during the pandemic. I joined the platform and started competing immediately. It gave me an opportunity to become involved in discussions and on the message boards, and the opportunity to compete on teams.”

Michiel George Ferreira (or MG) is a mathematician from South Africa. He holds a Bachelor’s Degree in Economics and an MSc in Mathematics from the University of Pretoria. His love for mathematics and desire to stay relevant with the latest development in the labour market at the time led him to further his studies in this field upon completion of his first degree. The switch from Economics to Mathematics allowed him to pursue a more technical career which continues to hold his fascination. Currently, he focuses on building models, developing software for business solutions, and providing technical assistance in his capacity as a freelancer.

The power of online communities

“There is so much information ready to be learned online, so I try to stay in touch with online learning possibilities,” MG says. “This is especially useful for the latest techniques in machine learning.”

During the Covid-19 pandemic and lockdown across the world, MG found himself in more online spaces, attending virtual meetings and conferences, becoming a frequent user on tech discussion platforms like StackOverflow, where he has since become a top contributor. While many of us struggled over the two years of pandemic and lockdown conditions, for MG this was a productive period as he got to connect with like-minded individuals.

One such connection was the online NVIDIA conference, held in April 2021, where he learned about Zindi during the session on AI in Africa. He joined Zindi afterward and, according to him, it has been an exciting experience competing and connecting with amazing talents.

“Zindi is a diverse community that gives everyone an opportunity to grow. Although there is a stark difference between being a casual member and being an active competitor on the platform, Zindi offers both users a space to learn and grow at their own pace.”

In his quest for continued growth, MG recently emerged as joint 1st place winner with another Zindi user. The Radiant Earth Spot the Crop XL Challenge tasked participants with using machine learning models and satellite time-series data to predict crop types in the Western Cape, South Africa.

“I would encourage everyone to compete — you gain invaluable practical experience and the competitive environment forces you to use the best techniques. At the same time, the score does not mean much and, while you should play to win, realise that the benefit is the journey and not the destination.”

The potential in competitions

In data science, there aren’t many ways to test out newfound knowledge that one might have learned. But with Zindi, users are able to gain practical implementation experience in competition spaces. Aside from the sense of community, this is one thing MG has come to love and enjoy about Zindi. He strongly believes that competing online is the best way to calibrate your knowledge so that what you have isn’t only theoretical but also of practical value.

“Competitions aren’t the real world. They’re not real-life events but they’re pretty close to solving real-world problems,” he points out. “They mimic real-life to sufficient levels for individuals to test their newly acquired skills by benchmarking their knowledge while developing their skills.”

MG is also passionate about the community and the incredible opportunity for growth the Zindi platform provides. He advises other members to participate actively on the platform — engaging on message boards, joining in discussions, and competing in teams rather than competing alone. He believes that engaging with others also improves an individual’s experience which is beneficial both ways.

For MG, machine learning modelling is very much a hobby as well as a day job that he hopes to build a sustainable business out of. In the future, he looks forward to working with other members on the Zindi platform while using the satellite mapping modelling that he has grown quite accustomed to.