Zindi: Climate and Health Prediction

This Zindi challenge asks you to build a supervised classifier that predicts whether a recorded death falls into a climate-sensitive category, using health, demographic, geographic and climate variables such as rainfall and temperature. The framing is explicitly about low-resource settings, so the problem is close to home.

Zindi has tagged it beginner friendly, which makes it a realistic first competitive project for a student who has finished an introductory machine learning course and wants a public leaderboard result to point at. It is open to all nationalities, Cameroon included.

You may enter alone or in a team of up to four, with 10 submissions a day and 300 in total — plenty of room to experiment.

The money is modest: USD 500, 300 and 200 for the top three. Treat this as portfolio-building rather than income.

The rule most likely to trip you up is the tooling restriction. Only open-source packages, only the data provided, no AutoML tools, and nothing that needs a credit card. As with all Zindi challenges the top ten are emailed at the close and have 48 hours to supply code that reproduces their score.

Zindi publishes the close date as 18 October 2026 with no clock time, so submit early.