Auto-PyTorch vs FEDOT

Auto-PyTorch is a free, self-hosted machine-learning toolkit based on PyTorch. It automates algorithm selection and hyperparameter tuning, including neural-network architecture and training settings, within a user-defined search budget. Documented search methods include Bayesian optimisation, meta-learning, SMAC and Hyperband. Supported work includes tabular classification, tabular regression and time-series forecasting. Tabular preprocessing can perform imputation, categorical encoding, scaling and feature processing, while related hyperparameters are tuned. The toolkit can build ensembles from validation-set predictions, with controls for ensemble size and candidate limits. Users can set estimator memory limits and total wall time. Parallel Bayesian optimisation uses Dask.distributed; workers need shared filesystem access to training data and models. Installation documentation specifies Linux, Python 3.7 or later, a C++11-capable compiler and SWIG 3.0. SWIG 4.0 or later is unsupported. Forecasting needs extra dependencies. A Docker image and PyPI installation are documented, under the 3-clause BSD license.

FEDOT is a free, open-source framework for automating the creation of data-driven composite models. It supports classification, regression, clustering and time-series forecasting, with tabular, text and image inputs, including multimodal combinations. Its workflow covers preprocessing, model selection, tuning, cross-validation and serialization. Users can leave parameters unspecified for fuller automation or provide them to guide pipeline composition. FEDOT uses the GOLEM library to optimize graph-based pipelines with meta-heuristic methods. Presets include best_quality, fast_train, stable, auto, gpu, ts and automl; auto is the default. Data can come from CSV files, pandas DataFrames, NumPy arrays or time-series CSV data. It can be installed with pip, and the API can be called from a console without Python code, saving predictions as CSV files. GPU evaluation uses RAPIDS and supports a specified set of models. The project is distributed under the BSD 3-Clause license and supports Windows, Linux and macOS.