Hello,
I've build a xgboost model and used lime to explain the predictions. The problem is my data is build so it contains null's. Xgboost can handle this quite good, the problem is that lime can't handle null's the same way. Is the any way to by pass this? someone that faced the same problem?
Hello, I've build a xgboost model and used lime to explain the predictions. The problem is my data is build so it contains null's. Xgboost can handle this quite good, the problem is that lime can't handle null's the same way. Is the any way to by pass this? someone that faced the same problem?