What is FEDOT

FEDOT is an open-source framework for automated modeling and machine learning (AutoML). It produces a lightweight end-to-end ML solution in an automated way using an evolutionary approach.

FEDOT logo

FEDOT supports classification (binary and multiclass), regression, and time series forecasting tasks. FEDOT works both on unimodal (only tabular/image/text data) and multimodal data (more than one data source).

Pipeline schema

FEDOT supports a full life-сyсle of machine learning task that includes preprocessing, model selection, tuning, cross validation and serialization.

model = Fedot(problem='classification', timeout=5, preset='best_quality', n_jobs=-1)
model.fit(features=x_train, target=y_train)
prediction = model.predict(features=x_test)
metrics = model.get_metrics(target=y_test)

Once FEDOT finds the best solution you have an opportunity to save it and look closer to the solution and optimization process if needed.

Example of solution

Framework uses ML models mostly from sklearn, statsmodels and keras libraries.

You can explore more concepts here.

There is a quickstart guide for beginers.