support transformation functions

dmitryikh / leaves

pure Go implementation of prediction part for GBRT (Gradient Boosting Regression Trees) models from popular frameworks

MIT License

419 stars 72 forks source link

Closed dmitryikh closed 5 years ago

dmitryikh commented 5 years ago

Currently leaves outputs prediction as a raw score. Client code should transform it to probabilities (logistic), lambdarank and so.

Let's introduce this ability to leaves.

Short list for XGBoost:

Short list for LightGBM: todo

dmitryikh commented 5 years ago

partially done in #54

but not all transformation functions were implemented. Currently here are logistic & softmax transformations for Lightgbm and XGBoost.

Close for now