hgboost’s documentation!

The Hyperoptimized Gradient Boosting library (hgboost), is a Python package for hyperparameter optimization for XGBoost, LightBoost, and CatBoost. HGBoost will carefully split the dataset into a train, test, and an independent validation set. Within the train-test set there is the inner loop for optimizing the hyperparameters using Bayesian optimization (based on Hyperopt) and, the outer loop is to test how well the best-performing models can generalize using an external k-fold cross validation. This approach will select the most robust model with the highest performance.

_images/schematic_overview.png

hgboost is fun because:

    1. It consists three of the most popular decision tree algorithms; XGBoost, LightBoost and Catboost.

    1. It consists the most popular hyperparameter optimization library for Bayesian Optimization; Hyperopt.

    1. An automated manner to split the data set into a train-test and independent validation to reliably determine the model performance.

    1. The pipeline has a nested scheme with an inner loop for hyperparameter optimization and an outer loop with k-fold crossvalidation to determine the most robust and best-performing model.

    1. It can handle both classification and regression tasks.

    1. It is easy to go wild and create a multi-class model or an ensemble of boosted decision tree models.

    1. It takes care of unbalanced datasets.

    1. It aims to create explainable results for the hyperparameter search-space, and model performance results by creating insightful plots.

    1. It is open-source.

    1. It is documented with many examples.

You contribution is important

If you ❤️ this project, star this repo at the github page and have a look at the sponser page!

Github

Please report bugs, issues and feature extensions at github.

Content

Indices and tables