# 12th solution

Congrats to winners especially for weiwei. And have a surprise that Faron didn't submit finally.
Thanks for sh1ng's baseline and Faron's F1 optimizer script.
sjv's DL solution is wonderful,thanks for sharing.
MODEL:
1.user product pair model.
2.if none model.
FEATURES:
User X Order features:
Products features:
Aisle features:
Department  features:
User features:
User X Products features:
User X Aisles features:
User X Department features:
Products X time features:
order_number features:
DATA:
I add the last prior as train data.Also,I add the second to the last prior,but not didn't improve much.
ENSEMBLE:
all["label"]=0.28*all["prediction_arbor"]+0.12*all["prediction_lgbm"]+0.36*all["label_2"]+0.24*all["label_1"]
prediction_arbor AND prediction_lgbm is from sh1ng's  baseline.
label_2 trained with shuffle, label_1 traind split by userid.They have the same features.
SUB:
Use faron's script with my "none model".
code link：
[https://github.com/plantsgo/Instacart-Market-Basket-Analysis/tree/master](https://github.com/plantsgo/Instacart-Market-Basket-Analysis/tree/master)
Thanks~and hope for more sharing~