Vlad Apetrei authored Jun 17, 2019 and David Monllaó committed Sep 09, 2019

Adds multi-class capabilities to prediction processors
as well as multi-classification unit tests

- Basic unit test for minimum machine learning backends requirements
- Warning return messages now include not enough data
- Clear models when the predictions processor is changed
- Refined the name of a couple of constants / methods

- Method names renamed to avoid interface changes once
  we support regression and unsupervised learning
- Adding regressor interface even if not implemente
- predictor interface comments expanded
- Differentiate model's required accuracy from predictions quality
- Add missing get_callback_boundary call
- Updated datasets' metadata to allow 3rd parties to code
  regressors themselves
- Add missing option to exception message
- Include target data into the dataset regardless of being a prediction
  dataset or a training dataset
- Explicit in_array and array_search non-strict calls
- Overwrite discrete should_be_displayed implementation with the binary one
- Overwrite no_teacher get_display_value as it would otherwise look
  wrong
- Other minor fixes

Part of MDL-57791 epic.

- Split model::predict in parts
- JS promises updated according to eslint-plugin-promise
- New API methods replacing direct DB queries
- Reduce insights nav link display cost
- Increase time limit as well as memory for big processes
- Move prediction action event to core
- Dataset write locking and others
- Refine last time range end time
- Removed dodgy splitting method id to int
- Replace admin_setting_predictor output_html overwrite for write_setting overwrite
- New APIs for access control
- Discard invalid samples also during prediction

Part of MDL-57791 epic.

Part of MDL-57791 epic.

Part of MDL-57791 epic.