GeoAI Ground-level NO2 Estimation Challenge by ITU
Helping Italy
1 000 CHF
Challenge completed ~1 year ago
Prediction
804 joined
372 active
Start
May 22, 24
Close
Nov 15, 24
Reveal
Nov 15, 24
'%3e%3ccircle r='120' fill='%230f47af'/%3e%3cg id='a'%3e%3cpath fill='%23fcdd09' d='m0-96-4.206 12.944 17.348 53.39h-23.13l-2.599 8h74.163l11.011-8H21.553z'/%3e%3cpath stroke='%23fcdd09' stroke-width='4' d='m25.863-35.597 30.564-42.069'/%3e%3c/g%3e%3cuse xlink:href='%23a' transform='rotate(72)'/%3e%3cuse xlink:href='%23a' transform='rotate(144)'/%3e%3cuse xlink:href='%23a' transform='rotate(216)'/%3e%3cuse xlink:href='%23a' transform='rotate(288)'/%3e%3c/g%3e%3c/svg%3e){kind=small}
Cv vs lb
Data · 11 Nov 2024, 10:52 · 8
Hi guys how you doing with this competition. My cv and lb is not correlated that much my cv is 9.8 and lb is 8.86 how is yours ? And did anyone used elevation data is it improve the score ? I tried many techniques that I used ariqo challenge but didn't work this time I don't know why I think trusting lb is more important than cv this time.
Same here ,
Even after a groupflod like CV my notebook scores don't match my lb score , At this stade everytime I want to test a new Idea I test it directly on the lb
Elevation data lead to worse score for me , so I don't use it currently
Yeah, I tried Groupfold, and it has not worked for me.
While to me both scores are related, the lb is only 20% of the test data. So I'm not sure if the lb actually matters that much.
i am eager to see private score, I think will have huge shakeup
They are strongly correlated in my case
are you using groupkfold or kfold
Not really. I have a suspicion about the GT_NO2 distribution of the test set. So i just evaluate the perfomance locally according to it
Got it, good luck