xgboost-distribution 0.1.1

XGBoost for probabilistic prediction.

xgboost-distribution

XGBoost for probabilistic prediction. Like NGBoost, but faster, and in the XGBoost scikit-learn API.

Installation

$ pip install --upgrade xgboost-distribution

Usage

XGBDistribution follows the XGBoost scikit-learn API, with an additional keyword in the constructor for specifying the distribution (see the documentation for a full list of available distributions):

from sklearn.datasets import load_boston
from sklearn.model_selection import train_test_split

from xgboost_distribution import XGBDistribution


data = load_boston()
X, y = data.data, data.target
X_train, X_test, y_train, y_test = train_test_split(X, y)

model = XGBDistribution(
    distribution="normal",
    n_estimators=500
)
model.fit(
    X_train, y_train,
    eval_set=[(X_test, y_test)],
    early_stopping_rounds=10
)

After fitting, we can predict the parameters of the distribution:

preds = model.predict(X_test)
mean, std = preds.loc, preds.scale

Note that this returned a namedtuple of numpy arrays for each parameter of the distribution (we use the scipy naming conventions, see e.g. scipy.stats.norm).

NGBoost performance comparison

XGBDistribution follows the method shown in the NGBoost library, using natural gradients to estimate the parameters of the distribution.

Below, we show a performance comparison of the NGBoost NGBRegressor and XGBDistribution models, using the Boston Housing dataset and a normal distribution (similar hyperparameters). We note that while the performance of the two models is essentially identical, XGBDistribution is 50x faster (timed on both fit and predict steps):

Note that the speed-up will decrease with dataset size, as it is ultimately limited by the natural gradient computation (via LAPACK gesv). However, with 1m rows of data XGBDistribution is still 10x faster than NGBRegressor.

Full XGBoost features

XGBDistribution offers the full set of XGBoost features available in the XGBoost scikit-learn API, allowing, for example, probabilistic regression with monotonic constraints:

Acknowledgements

This package would not exist without the excellent work from:

• NGBoost - Which demonstrated how gradient boosting with natural gradients can be used to estimate parameters of distributions. Much of the gradient calculations code were adapted from there.

• XGBoost - Which provides the gradient boosting algorithms used here, in particular the sklearn APIs were taken as a blue-print.

Note

This project has been set up using PyScaffold 4.0.1. For details and usage information on PyScaffold see https://pyscaffold.org/.