The Zimnat Insurance Assurance Challenge by #ZindiWeekendz
$300 USD
Predict when an insurance policy will lapse in Zimbabwe
314 data scientists enrolled, 97 on the leaderboard
InsurancePredictionStructured
Africa
22 May 12:00 (60 hours)
Predicting the past is easy...
published 22 May 2020, 07:54

This year has meant a big change to behaviour all over the world, which in turn has caused a mad scramble as companies try to re-train ML models to adapt to the new normal. It highlights a challenge with ML - if there's a shift in the data then models trained on old data stop working so well. Covid is the disruption of today, but I thought it's worth noting that my beloved country has had many disruptions, some of which will reflect in the data of this competition.

I won't list all the events, but over the period covered by this challenge we've had a regime change ("Definitely not a coup"TM), protests, fuel shortages, inflation, floods, drought... Definitely a challenge to forecast! Good luck all :)

thanks for giving us some context for the competition, especially from a native of the country in question ! good luck to you too ;)

This will be a really interesting challenge, and even more important, thanks to the serious context you have provided us @Johnowhitaker. Machine Learning algorithms are learning a pattern, and if global crisis as such, completely distort the pattern, they are bound to fail, and bring up their own challenge of remodelling the present. Fingers crossed !!!