Sergey

Operation repo refactor

4

This is a 🔨 code refactoring. ## Summary Рефактор тегов и пресетов: 1. Разделение. Теперь пресеты никак не пересекаются с тегами и служат лишь как апи. 2. Использование `Enum` везде,...

List of random problems that may occurs in tests. Tests that fall due to uncorrect parameters (#1203) 1. RANSAC. In some cases for some `residual_threshold` values pipeline throughs error: "there...

In all data splitters multi target in `multi_ts` data is cut. If cut is disabled then Fedot falls.

Each verification rule that should work with atomized models should be tested.

Some problems that have not been solved with refactoring. - [ ] https://github.com/aimclub/FEDOT/pull/1004/files#r1410530615 - [ ] Problems with nan encoding: - [ ] https://github.com/aimclub/FEDOT/pull/1004/files#r1410543794 - [ ] https://github.com/aimclub/FEDOT/pull/1004/files#r1410529619 - [...

Aggregated operations

4

There are some groups of models: 1. Linear (Lasso, Linear, Ridge, Elastic) 2. Boostings (XGBoost, CatBoost, LGBM, Sklearn boosting) 3. Forests (Random, ExtraTrees) 4. and so on Each group can...

Tests for correct conditions: 1. Classification tasks: 1. Binary classification 2. Multitarget classification 1. Regression task 2. Time series forecasting tasks: 1. Unidimensional case 2. Multi_ts case 3. Multidimensional time...