Value Selection Widget with Gradient Boosted Tree Learner

Hi everybody

you might now the dataset in-vehicle-coupon reccomendation.

In-Vehicle Coupon Recommendation - UCI Machine Learning Repository

I would like to create a machine learning model where you can choose parameters (variables) of different columns to get a forecast whether the coupon (with the chosen variables) will be accepted or not (prediction probability)

for the variable input I would use the node “Value Selection Widget”.

Variables would be destination, weather, passanger, gender, age, materialstatus, children, income, material status, income, education, occupation, income, coupon, bar, coffehouse, carryaway, restaurant<20, restaurant 20-50, tocoupongeq15min, tocoupongeq25min.

I already set up a workflow but i don’t know where to put the value selection widget exactly and how to configure the ML Learner and predictor to get a result based on my variables.

Maybe you could help me to complete the workflow or give me a better way so solve this question.

Thank you very much in advance!