Peter Shevcnenko
Peter Shevcnenko
Closes #437. May be merged later, if a major release is required for this change. Upd.: I addressed the backward compatibility issue, mentioned in #437, by doing such: - Defined...
Adds an option to save graph to OptHistory after each objective evaluation in the process of tuning (Solves #226). Allows tuners to work with Individual and Fitness for graphs comparison...
Tuner interface should be able to save tuning history or at least tuning results with estimated metrics to OptHistory.
As a part of #224 Plan: - Try to set seed for stochastic visualizations - Create reference outputs for deterministic visualizations with different combinations of parameters - Write tests, comparing...
- Some current meta-learning components include multiple transformations with reuse of other ones. - Moreover, there is excess usage of hardcoded configuration and data files affecting the components' behaviour. -...
- The usage of other components must be moved outside, their outputs should be used as input instead. These components are: `PymfeExtractor` and `FeaturePreprocessor` - The usage of saved files...
Текущие requirements приводят к [конфликту зависимостей](https://github.com/ITMO-NSS-team/GAMLET/actions/runs/6737255044/job/18314306021). @valer1435, назначаю тебя, потому что твои наработки используют FEDOT.Industrial. Укажи подходящий стабильный релиз в зависимостях.
We may add a similarity search that functions the following way: 1. Extract necessary metafeatures from a custom dataset (implement corresponding MetaFeaturesExtractor in #3 or in another PR) 2. Load...
Предлагаю здесь обсудить роли, форму и взаимодействие различных компонентов системы. Вопросы, предложения, всё приветствуется
As CodeCov reports, current Pytest coverage is around 42%. We should write unit tests to cover at least 80% of the framework. CodeCov reports should be used to define functionality...