SUA Outsmarting Outbreaks Challenge
Helping Tanzania, United Republic of
$12 500 USD + AWS credits
Completed (almost 2 years ago)
Prediction
817 joined
390 active
Start
06 Dec 24
Close
31 Jan 25
Reveal
01 Feb 25
 scale(.21174)'%3e%3cpath d='m52.487 177.086 13.018 11.852 11.762-11.762 11.896 12.266 12.846-11.358 11.827 12.43 13.339-11.013 10.897 12.728 13.27-10.145 13.947 12.319'/%3e%3ccircle cx='89.425' cy='154.584' r='5.167'/%3e%3ccircle cx='111.237' cy='156.047' r='5.167'/%3e%3ccircle cx='30.05' cy='16.5' r='4'/%3e%3c/g%3e%3c/svg%3e)
Cholera
Data · 19 Jan 2025, 14:18 · 10
has anyone noticed the unusualness of the cholera instances
Yes, almost no rows in training data but a lot in test data. The same goes for some places and locations.
In other words, for some rows, we are supposed to make predictions for an unknown disease and an unknown place ! I'm puzzled about the train/test split.
Yea that's very confusing, how are we going to train the model for predicting cholera if we can't train the model with cholera instances
I think the test set gives you hints to handle the cholera classs
Thanks for the insight
You can get a hint with the test set. The tricky part is that the public leaderboard seems to not consider the cholera rows
You mean to say the public leaderboard is not evaluating the correctness of the cholera instances
Yes
I won't worry too much about cholera as no model will be able to predict out-of-distribution values very well.
Yeah that's true, the dataset is heavily skewed
Thank you for highlighting this. I think it's either there was an outbreak of cholera or the reporting regulations changed and every facility was supposed to inlude cholera in their monthly reports.