How to interpret the exel file and use it to improve the modelisation

[bourbe]

Hi,

I recently discover this tool and thank you for the port in python. I have' some questions:

• i understand very well the first sheet and I see it is a kind of feature importance by each score
• we have the same principle in sheet 2 for interaction but first what you understand by interraction here ?
• for example in the iris data, the interaction petal length (cm)|petal width (cm) has the best gain score. How to use this information in the modlisation ?

thanks you for your answer

[JoshuaC3]

I was also wondering what the Leaf Statistics show? What information can you glean about the model from these statistics?