Work on combining policy trees and decision trees into singular form

[laDok8]

Summary of Changes:

• Self-loop Action Addition: Updated models in models/archive/atva24-policy-trees to include the _stay_ action. (Note: Some benchmarks, such as virus and uav, have not yet been updated.)
• New CLI Flag: Introduced the --ldokoupi flag to the CLI, enabling preparation steps for combining trees. Default behavior remains unchanged.
• Policy Tree Merging: Enhanced the policy tree merging algorithm, resulting in fewer overall policies (depending on the benchmark) while improving robustness.
• Policy reductions Post-processed policies derived from stochastic games to reduce their size by minimizing state-action pairs.
• Refactoring: Various code improvements and cleanup.