Radiant Earth Spot the Crop XL Challenge
'%3e%3cpath fill='%23001489' d='M0 0v6h9V0z'/%3e%3cpath fill='%23e03c31' d='M0 0v3h9V0z'/%3e%3cg stroke='%23fff' stroke-width='2'%3e%3cpath id='d' d='m0 0 4.5 3L0 6m4.5-3H9'/%3e%3cuse xlink:href='%23b' stroke='%23ffb81c' clip-path='url(%23c)'/%3e%3c/g%3e%3cuse xlink:href='%23d' fill='none' stroke='%23007749' stroke-width='1.2'/%3e%3c/g%3e%3c/svg%3e)
Helping South Africa
$8 800 USD
Challenge completed ~4 years ago
Classification
Earth Observation
347 joined
50 active
Start
Jul 05, 21
Close
Oct 03, 21
Reveal
Oct 03, 21
Top 2 and 3 solution!
Notebooks Ā· 6 Oct 2021, 12:13 Ā· 5
Hi, Zindi!
I was very glad to participate in this wonderful competition. Thank you very much for organizing it!
Thanks to the participants who made so much serious competition and made me work hard 😅
I posted my solution on github: https://github.com/sweetlhare/Radiant-Solution
If there are any questions, write in this thread, I will answer. The main thing that is in the solution: converting images into plates and calculating vegetation indices, various engineering features and the use of geo-signs, the use of CatBoostClassifier and powerful LightAutoML.
If my solution was useful for you, put a ⭐️!
From Russia with love,
Your DS
P.S. now I'm thinking of traveling to Africa, I liked participating with you so much!)
This is a good solution, but Automl algos are not permitted as per the Zindi Rules.
This is a strange rule, because LightAutoML is the most common stacking of catboost and lgbm models. What's the point of banning it?
@Zindi What do you think about it? LightAutoML is just a tool for faster stacking and blending of catboost and lgbm models
The solution must use publicly-available, open-source packages only. You may use only the datasets provided for this competition. Automated machine learning tools such as automl are not permitted.
That's clear in the rules of the competition
I didn't sweep all the solution and maybe they could be something else wrong in your solution because again geofeatures like x,y are not permitted in this competition and they were clear about it also.
I hope Zindi make a clear decision about this.
@ZINDI
The host was clear in the webinar, related geofeatures like what is the nearest crop (label) or what's the other farmers plant around it are not allowed. 'Which looks many players used those information" @Zindi