Implement automated model selection and architecture search

[gelluisaac]

Description

    Build an automated model selection system that uses neural architecture search and meta-learning to recommend optimal model architectures.

    ## Files to modify
    - astroml/training/model_selection/automl.py
    - astroml/training/model_selection/nas.py
    - astroml/training/model_selection/meta_learning.py
    - astroml/training/model_selection/benchmark.py
    - astroml/api/routers/model_selection.py

    ## Out of scope
    - Refactoring unrelated modules
    - Changing public API contracts
    - Major version bumps or breaking changes
    - Performance optimizations not directly related to the issue

    ## Acceptance criteria
    - [ ] Code changes implemented as per procedure steps
    - [ ] All new code has complete type hints
    - [ ] Unit tests added/updated with >90% coverage for changed code
    - [ ] All tests pass: pytest tests/
    - [ ] Linting passes: pre-commit run --all-files
    - [ ] Documentation updated if needed
    - [ ] Changes reviewed and approved

    ## Procedure
    1. Implement AutoML model selection pipeline
    2. Build neural architecture search (NAS) framework
    3. Add meta-learning for task similarity
    4. Implement model benchmarking suite
    5. Create model recommendation engine
    6. Add computational budget constraints
    7. Implement model selection API
    8. Test with various ML tasks