Mlrl-boomer

A bugfix release that comes with the following changes.

Fixes

- Fixed faulty calculation of gradients and Hessians when using the BOOMER regression algorithm with the command line arguments `--loss squared-error-non-decomposable` and `--head-type single`.

Quality-of-Life Improvements

- Linux ARM64 packages are now build natively instead of using virtualization, which speeds up the build process tremendously.
- Continuous Integration is now used to automatically update the versions of GitHub runners on a regular basis.

A bugfix release that comes with the following changes.

Fixes

- When building macOS packages, we do now manually compile OpenMP to ensure that the bundled library matches the target platform.

Quality-of-Life Improvements

- The build system now uses a lightweight custom implementation instead of [SCons](https://scons.org/) and is better modularized to avoid unnecessary runs of Continuous Integration jobs when only certain parts of it are modified.
- Releases are now automated via Continuous Integration, including the update of the project's changelog.
- The presentation of algorithmic parameters in the documentation has been improved.
- Continuous Integration is now used to automatically update outdated GitHub Actions, outdated Python dependencies, and the Doxygen configuration on a regular basis.
- Continuous Integration is now used to check the documentation for broken links and spelling mistakes.
- The syntax of TOML files is now validated by applying [taplo](https://github.com/tamasfe/taplo) via Continuous Integration.

A bugfix release that fixes the following issues:

- A crash has been fixed that could occur when using the command line argument `--print-rules` or `--store-rules` with a dataset that contains nominal features.

A major update to the BOOMER algorithm that introduces the following changes.

{warning}
This release comes with several API changes. For an updated overview of the available parameters and command line arguments, please refer to the [documentation](https://mlrl-boomer.readthedocs.io/en/0.11.0/).


Algorithmic Enhancements

- **The BOOMER algorithm can be used for solving regression problems**, including single- and multi-output regression problems.

Additions to the Command Line API

- **Custom algorithms can now be easily integrated** with the command line API due to the ability to dynamically load code from your own Python modules or source files, as illustrated [here](https://mlrl-boomer.readthedocs.io/en/0.11.0/user_guide/testbed/runnables.html)
- **The value to be used for sparse elements in the feature matrix can be specified** via the argument `--sparse-feature-value`.

API Changes

- The Python module or source file providing an integration with the machine learning algorithm to be used by the command line API must now be specified as described [here](https://mlrl-boomer.readthedocs.io/en/0.11.0/user_guide/testbed/arguments.html#basic-usage).
- Several parameters and their values have been renamed to better reflect the scope of the project, which now includes multi-output regression problems. For an up-to-date list of parameters, please refer to the [documentation](https://mlrl-boomer.readthedocs.io/en/0.11.0/).
- Rules with complete heads are now learned by default when using a decomposable loss function and a dense format for storing statistics.

A bugfix release that fixes the following issues.

- A rare issue that may result in rules with empty heads being learned when using the argument `--head-type partial-dynamic` has been fixed.

A bugfix release that comes with the following changes.

- If the sparse value of a feature matrix is provided to the Python API, it is now properly taken into account when converting into a dense matrix.
- The C++ code is now checked for common issues by applying `cpplint` via Continuous Integration.
- The styling of YAML files is now verified by applying `yamlfix` via Continuous Integration.