What has changed

Added initialize parameter to all distribution classes, allowing the user to choose if distributional parameters are to be initialized with unconditional start values before model training. When enabled (initialize=True), this can help improve speed in some cases, though it may lead to early stopping or suboptimal solutions if the unconditional start values are far from optimal. The parameter defaults to False since users have reported too early stopping during cross-validation using the previously default initialization.

General

We appreciate the valuable feedback and contributions from our users, which have helped us in making LightGBMLSS even better. We encourage you to update to this latest version to take advantage of the new features and improvements. To get started, check out the documentation and examples.

Thank you for your continued support, and we look forward to your feedback.

Happy modeling!