choshin84 / learning_memo

personal learning memo
0 stars 0 forks source link

Model evaluation: random vs. explicit split #17

Open choshin84 opened 5 years ago

choshin84 commented 5 years ago

Tweek summary

when random split, i.e. k-fold CV, it's testing model over-fit, vs. explicit split, i.e. based on time frame, it's testing model prediction capability.