For time-series forecasting models, random K-fold is dangerous beca…

Question

For time-series forecasting models, random K-fold is dangerous because:

Accepted Answer

It can leak future into past training folds; use time-aware splits — Time-series needs time-aware splits — random K-fold leaks the future.

Answer

It is always perfect

Answer

Clocks do not exist

Answer

Delta cannot store timestamps

For time-series forecasting models, random K-fold is dangerous because: