Add abstract meta learning approach

[MorrisNein]

This class is meant to combine meta-learning components into templates to simplify their application.

As an example, I added the implementation of KNNSimilarityModelAdvice that I use in experiments.

TODO:

• implement unit tests
• add docstrings

[pep8speaks]

Hello @MorrisNein! Thanks for updating this PR. We checked the lines you've touched for PEP 8 issues, and found:

• In the file meta_automl/approaches/__init__.py:

Line 1:1: F401 '.meta_learning_approach.MetaLearningApproach' imported but unused

Comment last updated at 2023-12-19 10:16:53 UTC

[codecov]

Codecov Report

Attention: 130 lines in your changes are missing coverage. Please review.

Comparison is base (5c11fb3) 30.29% compared to head (1901863) 28.71%.

❗ Current head 1901863 differs from pull request most recent head 4ff939c. Consider uploading reports for the commit 4ff939c to get more accurate results

Files	Patch %	Lines
...a_automl/approaches/knn_similarity_model_advice.py	0.00%	76 Missing ⚠️
...preparation/models_loaders/fedot_history_loader.py	0.00%	37 Missing ⚠️
meta_automl/approaches/meta_learning_approach.py	0.00%	15 Missing ⚠️
meta_automl/approaches/__init__.py	0.00%	1 Missing ⚠️
.../meta_features_extractors/dataset_meta_features.py	0.00%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #88      +/-   ##
==========================================
- Coverage   30.29%   28.71%   -1.59%     
==========================================
  Files          56       60       +4     
  Lines        2340     2469     +129     
==========================================
  Hits          709      709              
- Misses       1631     1760     +129     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[MorrisNein]

@pep8speaks suggest diff