A single ML model for all?

Water Supply Forecast Rodeo
1

Hello,

Maybe I’m missing something here, but the Problem Description doc does not dictate whether we should train a single ML model to predict at all Sites or we can train a separate ML model for each and different site depending on feature availability? Please clarify. Thank you.

2

Hi @dialuser88,

The competition does not impose any requirements about whether you use a single model or many models—it’s up to you to decide.