Zindi will host UmojaHack Africa, the first-ever pan-African inter-university machine learning hackathon, virtually due to coronavirus outbreak.
The UmojaHack Africa competition which takes place on Saturday, 21 March 2020.
Zindi will host the first-ever virtual inter-university artificial intelligence (AI) hackathon for students across Africa. Over 70 universities from 17 African countries, with an estimated 2500 students, have registered to participate in the hackathon taking place on Zindi’s online platform.
“This level of collaboration and partnership across Africa is very exciting, and it is quite simple for students to join remotely through Zindi. The day’s activities will be live-streamed on our YouTube channel, from the launch of the hackathon through to when we announce the winners at the end of the day,” Celina Lee, Zindi’s chief executive officer says.
“We will host discussions about the three challenges, check-in with universities and participating students around Africa, and respond to questions via live streaming. Zindi’s platform makes this all possible despite the health challenges the world is currently facing.”
www.zindi.africa is a free-to-use data science competition platform which requires participants to use universally compatible open-source languages and tools only, making it widely accessible to communities and organizations around Africa. Students at each university will work in teams of up to four people, competing for cash and in-kind prizes while representing their university. The stakes are high and Zindi expects the best of the next generation of data scientists to show up on the day.
The top-performing team in each of the three challenges will be awarded prize money of $1000, $600, and $400 respectively. Teams placing 2nd and 3rd in each category will be awarded learning opportunities from one of the sponsors of the event. The associated university will receive $7500 (1st category), $5000 (2nd category), and $2500 (3rd category).
These teams of data scientists will be working to find the best solution for three real-world challenges:
- Hotspot Fire Prediction Challenge
Teams are challenged to figure out the dynamics of where and when fires occur in the Democratic Republic of Congo, a country dealing with the health and environmental impacts of slash and burn agriculture. They must also predict how these dynamics will play out in the future, under different climatic conditions. The challenge is to build a model capable of predicting the burned area in different locations.
- SAEON Marine Invertebrate Classification Challenge
Teams are challenged to develop an automated image classification solution for photographs of marine invertebrate taken by researchers in South Africa. The solutions will reduce the resources needed to manually process photos and allow researchers to identify changes in marine life faster and more efficiently, to better monitor and understanding climate change impacts.
- Xente Purchase Prediction Challenge
Teams are challenged to create a machine learning model to predict purchase behaviour of customers on the Xente app, based on their purchase history. Xente is a Ugandan e-commerce start-up that makes it easy for consumers to make payments, get loans, and shop using simply a mobile phone. The resulting models and solutions will enable target marketing and create a better customers experience.
The machine learning solutions developed during UmojaHack Africa and the skills developed by the participants promise large-scale improvements to people’s lives, and it is no surprise that some top brands have got in on the action and partnered with Zindi to sponsor this exciting venture. Sponsors of the event include Microsoft, African Bank, Google Research, Liquid Telecom, Alliance4ai, Instadeep, and the Field Institute.