How did you achieve an RMSE of 11? @machine_learning My team and I have been working on this data for a long time, but we can't reach 11. Could you share some insights if you don't mind?
As I learnt from my kaggle experience : you choose a risky sub like those using post processing and safe submission based on CV. (for me I have 12.xx submission that i trust the most)
I drew inspiration from top solutions of this kaggle competition:https://www.kaggle.com/competitions/playground-series-s3e20/overview.
I used a kind on risky "POST PREPROCESSING" like this top solution https://www.kaggle.com/competitions/playground-series-s3e20/discussion/433567.
As I learnt from my kaggle experience : you choose a risky sub like those using post processing and safe submission based on CV. (for me I have 12.xx submission that i trust the most)
Alright thank you @machine_learning for the insight.