I’d just like to know if it’s possible to share the Test Data - Household Consumption Labels and Test Data - Poverty Rate Labels. I have a bunch of models that I left out of the competition to avoid falling too far behind the public leaderboard.
Thank you for the competition. It was a great exercise for me after seeing someone’s post on LinkedIn. I told myself I could do better, and I did, much better. Now I’d like to know if my attempts that didn’t make the cut could have done better in the competition. That’s all.
I’d also love to see the test labels so I can learn from them about how to create better generalising models in a more principled manner. I’m guessing the private leaderboard was 1 or 2 of the test surveys, and I’d very much appreciate being able to see in more detail why my solution did not generalise better than it did. This was my first attempt at this kind of competition, and it was an interesting experience.