FEDOT

FEDOT is an open-source framework for automatically generating data-driven composite models. It supports classification, regression, clustering and time-series forecasting, and works with tables, text, images and combinations of these data types. Its workflow covers preprocessing, model selection, tuning, cross-validation and serialization. Users can leave parameters out for fuller automation or supply them to guide pipeline composition. FEDOT uses the GOLEM library to optimize graph-based pipelines with meta-heuristic methods, and offers presets including best_quality, fast_train, stable, auto, gpu, ts and automl. Inputs can come from CSV files, pandas DataFrames, NumPy arrays or time-series CSV data. Its API can also be called from a console without Python code, saving predictions as CSV files. GPU evaluation uses RAPIDS and supports a specified set of models, including Ridge, Lasso, LogisticRegression, random forest classifiers and regressors, KMeans and SVC. Installation is available with pip, with optional image, text-processing and DNN dependencies. FEDOT is free under the BSD 3-Clause license and runs on Windows, Linux and macOS.

FEDOT covers several machine-learning tasks and data types, with adjustable automation and a console option. Users considering GPU evaluation should check that the listed model support fits their needs.

FEDOT may suit developers and researchers who want to automate parts of model creation across supported data types and tasks. It also offers controls for users who prefer to guide pipeline composition.

Overview

FEDOT is a code-first AutoML framework for building composite machine-learning models, aimed at developers and researchers who want automation they can steer.

Key features

FEDOT supports binary and multiclass classification, regression, clustering, and univariate or multivariate time-series forecasting. It accepts tables, text, images, and multimodal combinations, with inputs from CSV files, pandas DataFrames, NumPy arrays, and time-series CSV data. That range suits projects with varied data sources, though it does not replace choosing suitable inputs and evaluation criteria.

The framework covers preprocessing, model selection, tuning, cross-validation, and serialization. Its preprocessing handles infinite and missing values, binary and non-binary categorical features, and extra spaces in categorical data, reducing routine cleanup. The default cross-validation setting is five folds; adding metrics to the optimizer can help address potential bias.

Automation can be full when parameters are omitted, or partial when users supply parameters and compose pipelines manually. Presets include best_quality, fast_train, stable, auto, gpu, ts, and automl, with auto as the default. This range supports quick starts as well as more directed work, but making the most of the controls calls for code comfort.

FEDOT uses GOLEM to optimize and learn graph-based pipelines with meta-heuristic methods. Its models come mostly from scikit-learn, statsmodels, and Keras; the project also supports scikit-learn, CatBoost, and XGBoost, and allows custom libraries to be integrated. GPU evaluation uses RAPIDS and supports Ridge, Lasso, LogisticRegression, RandomForestClassifier, RandomForestRegressor, KMeans, and SVC. That GPU option is useful for matching work to supported estimators, but it is not broad GPU coverage across every model.

Installation is available through pip install fedot; optional image, text-processing, and DNN dependencies are available through fedot[extra]. The API can also be called from a console without Python code, with predictions saved as CSV files. This widens access beyond writing Python scripts, while the workflow remains code-oriented rather than graphical.

Pricing

FEDOT is free and open source, with a free plan under the BSD 3-Clause license. There is no paid tier or trial to weigh against it. The free plan has no stated seat or quota cap in its terms, making it suitable for project and research use without a subscription commitment.

Platforms

FEDOT supports Windows, Linux, and macOS, and can be used through its API or in a self-hosted setup. The quick-start guide lists these operating systems, making it a practical fit for users who can run and manage software in their own environment.

Who it's for

FEDOT is best for developers and researchers who want an open-source AutoML framework across several data types and tasks, with the option to guide pipeline construction rather than accept full automation. Its BSD-3 license permits use in projects and research. It is a weaker fit for readers who need a graphical workflow or GPU evaluation for models outside the supported list.

Pros and cons

• Pros: Broad coverage of classification, regression, clustering, and time-series forecasting makes it applicable to varied modeling problems.
• Pros: Adjustable automation, presets, and extensibility let users balance a fast automated start with custom pipeline control.
• Pros: Multimodal inputs and preprocessing for common missing, infinite, and categorical-data issues reduce some setup work.
• Cons: The workflow is code-based, so it is not the right choice for users who need a graphical interface.
• Cons: GPU evaluation is limited to seven named estimators, which can rule it out for GPU-focused work using other models.

Alternatives

AutoML Software is a useful starting point for comparing options across the category. Choose Auto-PyTorch instead for a free, self-hosted Linux option developed by the AutoML Groups of the University of Freiburg and Hannover. AutoGluon is another free, self-hosted Python library, while AutoKeras is a free Python package installed with pip. Choose LightAutoML for a free open-source library installable from PyPI under Apache 2.0. BigML is an alternative with a free tier capped at 16 MB per task, two parallel tasks, and one user. JADBio offers a free Basic plan with one seat, three projects, a 50 MB upload limit, 500 MB storage, and one model export. Amazon SageMaker Autopilot is a paid, pay-as-you-go option with a free trial and no minimum fees or upfront commitments. Siemens Building X offers applications or APIs on yearly subscriptions with quote-based pricing.

Verdict

Choose FEDOT if you want free, open-source AutoML with broad task and data support, plus control over how much pipeline construction is automated. Its strongest reason to choose it is that combination of breadth and adjustability; look elsewhere if you need a graphical workflow or GPU support beyond its named estimators.