'XGBoost Regression - Possible to output vector rather than number at leaf node?
I'm using XGBoost in python to address a regression problem and it's working extremely well.
The application relates to physical layers in a system. We're currently estimating the top (physical) layer but we'd like to extend this so rather than produce 1 target value, it produces a target vector which can represent the top x layers in the system (where x is relatively small, say 5 layers).
We have all the data to train this but can XGBoost produce a vector/list of outputs rather than a single numerical value at a leaf node?
Sources
This article follows the attribution requirements of Stack Overflow and is licensed under CC BY-SA 3.0.
Source: Stack Overflow
| Solution | Source |
|---|