AutoGluon-Tabular: Robust and Accurate AutoML for Structured Data

TLDR

Existing AutoML frameworks mainly focus on model and hyperparameter selection. The paper introduces AutoGluon‑Tabular, an open‑source AutoML framework that trains accurate models on raw tabular data with a single line of Python. AutoGluon‑Tabular builds multi‑layer ensembles of diverse models, stacking them, and is implemented as an open‑source framework requiring only a single line of Python, with its performance benchmarked against other AutoML platforms. Experiments on 50 Kaggle and OpenML tasks show AutoGluon is faster, more robust, and more accurate than competitors, often outperforming the best‑in‑hindsight combinations, and it beat 99 % of participants in two Kaggle competitions after 4 h of training.

Abstract

We introduce AutoGluon-Tabular, an open-source AutoML framework that requires only a single line of Python to train highly accurate machine learning models on an unprocessed tabular dataset such as a CSV file. Unlike existing AutoML frameworks that primarily focus on model/hyperparameter selection, AutoGluon-Tabular succeeds by ensembling multiple models and stacking them in multiple layers. Experiments reveal that our multi-layer combination of many models offers better use of allocated training time than seeking out the best. A second contribution is an extensive evaluation of public and commercial AutoML platforms including TPOT, H2O, AutoWEKA, auto-sklearn, AutoGluon, and Google AutoML Tables. Tests on a suite of 50 classification and regression tasks from Kaggle and the OpenML AutoML Benchmark reveal that AutoGluon is faster, more robust, and much more accurate. We find that AutoGluon often even outperforms the best-in-hindsight combination of all of its competitors. In two popular Kaggle competitions, AutoGluon beat 99% of the participating data scientists after merely 4h of training on the raw data.