maybe someone can help me with my problem. In an industrial farm 10 lines produce an certain amount of finished goods and a certain amount of defective products. The defects are categorized in about 20 types. I have collected for each month in 2024 the total amount of good products and the amount of products with defects. Here is an example for August 24
The problemi is that the amount output of the predictor do not depends on the previous table and the result is always the same no matter the values in the second table. Where is the mistake?
if you look at the tree (open view on the simple regression tree learner), you will see that the model only uses Month/year to build up the decision tree. Looking at the predicted value (10,160) in the tree, you can see which decisions lead there. Basically, this is the path it takes when it doesn’t know the month. In the table you want to predict, the value for Month/year is Giugno 2025, which the model has never seen before, so it will give you the default of 10,160.
