Zindi
CompeteJobsLearnChatLeaderboardFor Organisations
PartnersMeet the teamPressCase studiesAI4EAC
Primary competition visual

INEGI UN-GGIM Human Settlement Detection Challenge by ITU

Helping Mexico
1 000 CHF
Completed (~2 years ago)
Classification
213 joined
84 active
InfoDataChatLeaderboard
Starti
02 Sep 24
Closei
28 Sep 24
Reveali
10 Oct 24
User avatar
marching_learning
Nostalgic Mathematics
Let'share CV vs Leaderboard
Help Ā· 20 Sep 2024, 15:24 Ā· 8

Hello Everyone, I hope you do enjoy the competition.

I'm a bit curious about the correlation between CV (Cross Validation Score) and LB (leaderboard). I hope you'll be willing to share.

For me CV = 0.89 & LB=0.81.

Discussion 8 answers
User avatar
crossentropy
20 Sep 2024, 15:25

CV-> 0.83, LB -> 0.82

Upvotes 2
User avatar
Sodiq_Babawale_
20 Sep 2024, 15:47

CV = 0.85 LB = 0.82

Upvotes 2
User avatar
marching_learning
20 Sep 2024, 15:59

Wow thank you guys for your quick answer

Upvotes 1
User avatar
Acalo
20 Sep 2024, 18:00

This gbt or deep learning?

Upvotes 0
User avatar
Sodiq_Babawale_
21 Sep 2024, 19:03replied to Acalo

Boosting models

Upvotes 1
User avatar
nymfree
23 Sep 2024, 11:33

CV=0.88, LB=0.84. There's a real shake up possibility in this competition

Upvotes 1
User avatar
marching_learning
23 Sep 2024, 12:23replied to nymfree

If you don't mind @nymfree are you ubsing boosting or neural nets. I have the biggest gap between CV and LB. And I'm scratching my head to understand why ?

Upvotes 0
User avatar
nymfree
23 Sep 2024, 14:19replied to marching_learning

using neural nets. Haven't tried boosting yet.

Upvotes 1

Join the largest network for
data scientists and AI builders

AboutFAQsPrivacy PolicyTerms of UseRules
Status
Ā© Zindi 2026