'using H2O flow XGboost model
It gives a regression prediction as continuous score with negative values, like -1.27544 < x < 6.68112. How I interpret the negatives?
Solution 1:[1]
If you are using an H2O algorithm to predict a binary target (0/1), unless you convert your target column to a factor using (.asfactor() in python or as.factor() in R), H2O will assume this column is numeric and will solve a regression problem.
please verify the data type of your target column (it will likely show integer) and make sure that it shows enum.
more informations about your target distribution choices can be found here: http://docs.h2o.ai/h2o/latest-stable/h2o-docs/data-science/algo-params/distribution.html
Sources
This article follows the attribution requirements of Stack Overflow and is licensed under CC BY-SA 3.0.
Source: Stack Overflow
| Solution | Source |
|---|---|
| Solution 1 | Lauren |