This page contains a moderated list of examples, tutorials, articles, and research papers about AutoGluon use cases. It is inspired by awesome-machine-learning.
We will be happy to add your success story using AutoGluon to this list. Send us a pull request if you want to include your case here.
To get started, we recommend watching AutoGluon 1.0: Shattering the AutoML Ceiling with Zero Lines of Code, our talk at AutoML Conf 2023.
- AutoGluon-TimeSeries: Every Time Series Forecasting Model In One Library (Towards Data Science, Jan 2024)
- AutoGluon-TimeSeries: Creating Powerful Ensemble Forecasts - Complete Tutorial (AI Horizon Forecast, Dec 2023)
- AutoGluon for tabular data: 3 lines of code to achieve top 1% in Kaggle competitions (AWS Open Source Blog, Mar 2020)
- AutoGluon overview & example applications (Towards Data Science, Dec 2019)
AutoGluon is widely adopted on ML competition sites such as Kaggle. Below is a sampling of competition solutions that use AutoGluon to achieve strong results.
| Placement | Competition Solution | Author | Date | AutoGluon Details | Notes |
|---|---|---|---|---|---|
| 🥈 Rank 2/2392 (Top 0.1%) | Regression with an Insurance Dataset | SCRIPTCHEF | 2024/12/31 | v1.2, Tabular | Kaggle Playground Series S4E12. Also used in 9th and 10th place solutions! |
| 🥇 Rank 1/2687 | Exploring Mental Health Data | Mahdi Ravaghi | 2024/11/30 | v1.1, Tabular | Kaggle Playground Series S4E11. Also used in 4th and 13th place solutions! |
| Rank 8/3859 (Top 0.3%) | Loan Approval Prediction | Mahdi Ravaghi | 2024/10/31 | v1.1, Tabular | Kaggle Playground Series S4E10 |
| 🥇 Rank 1/3066 | Regression of Used Car Prices | Mart Preusse | 2024/09/30 | v1.1, Tabular | Kaggle Playground Series S4E9. Also used in 🥈 2nd, 🥉 3rd, 4th, and 5th place solutions! |
| 🥇 Rank 1/1116 | Kaggle AutoML Grand Prix (Overall) | Alexander R., Dmitry S., Rinchin | 2024/09/01 | v1.1, Tabular | Teams using AutoGluon in the Grand Prix: 🥇 1st, 🥈 2nd, 🥉 3rd, 4th, 6th, 7th, 8th, 9th, and 10th place teams! |
| 🥈 Rank 2/247 (Top 1%) | Kaggle AutoML Grand Prix Episode 5 | Robert Hatch | 2024/09/01 | v1.1, Tabular | Also used in 🥉 3rd, 4th, 6th, 7th, 9th, and 10th place solutions! |
| 🥇 Rank 1/2424 | Binary Prediction of Poisonous Mushrooms | Optimistix | 2024/08/31 | v1.1, Tabular | Kaggle Playground Series S4E8. Also used in 🥈 2nd, 🥉 3rd, 4th, 6th, 8th, and 10th place solutions! |
| 🥇 Rank 1/218 | Kaggle AutoML Grand Prix Episode 4 | Lennart P., Nick E. & Arjun K. | 2024/08/01 | v1.1, Tabular | Also used in 🥈 2nd, 🥉 3rd, 4th, 5th, 6th, 7th, 8th, 9th, and 10th place solutions! |
| 🥉 Rank 3/2236 (Top 0.2%) | Binary Classification of Insurance Cross Selling | Tilii | 2024/07/31 | v1.1, Tabular | Kaggle Playground Series S4E7 |
| Rank 4/207 (Top 2%) | Kaggle AutoML Grand Prix Episode 3 | Lennart Purucker & Nick Erickson | 2024/07/01 | v1.1, Tabular | |
| Rank 17/2684 (Top 1%) | Classification with an Academic Success Dataset | Mart Preusse | 2024/06/30 | v1.1, Tabular | Kaggle Playground Series S4E6 |
| 🥉 Rank 3/542 (Top 0.6%) | WiDS Datathon 2024 Challenge #2 | olgaskv | 2024/06/11 | v1.1, Tabular | |
| 🥇 Rank 1/230 | Kaggle AutoML Grand Prix Episode 2 | Lennart Purucker & Nick Erickson | 2024/06/01 | v1.1, Tabular | Also used in 5th place solution! |
| 🥇 Rank 1/2788 | Regression with a Flood Prediction Dataset | Alexandre Daubas | 2024/05/31 | v1.1, Tabular | Kaggle Playground Series S4E5. Also used in 🥈 2nd, 🥉 3rd, and 4th place solutions! |
| Rank 5/214 (Top 3%) | Kaggle AutoML Grand Prix Episode 1 | James King | 2024/05/01 | v1.1, Tabular | Also used in 8th and 9th place solutions! |
| 🥇 Rank 1/2606 | Regression with an Abalone Dataset | Johannes Heller | 2024/04/30 | v1.0, Tabular | Kaggle Playground Series S4E4. Also used in 🥈 2nd, 🥉 3rd, 4th, and 8th place solutions! |
| 🥉 Rank 3/2303 (Top 0.2%) | Steel Plate Defect Prediction | Samvel Kocharyan | 2024/03/31 | v1.0, Tabular | Kaggle Playground Series S4E3 |
| 🥈 Rank 2/93 (Top 2%) | Prediction Interval Competition I: Birth Weight | Oleksandr Shchur | 2024/03/21 | v1.0, Tabular | |
| 🥈 Rank 2/1542 (Top 0.2%) | WiDS Datathon 2024 Challenge #1 | lazy_panda | 2024/03/01 | v1.0, Tabular | |
| 🥈 Rank 2/3746 (Top 0.1%) | Multi-Class Prediction of Obesity Risk | Kirderf | 2024/02/29 | v1.0, Tabular | Kaggle Playground Series S4E2 |
| 🥈 Rank 2/3777 (Top 0.1%) | Binary Classification with a Bank Churn Dataset | lukaszl | 2024/01/31 | v1.0, Tabular | Kaggle Playground Series S4E1 |
| Placement | Competition Solution | Author | Date | AutoGluon Details | Notes |
|---|---|---|---|---|---|
| Rank 4/1718 (Top 0.2%) | Multi-Class Prediction of Cirrhosis Outcomes | Kirderf | 2023/12/31 | v1.0, Tabular | Kaggle Playground Series S3E26 |
| 🥈 Rank 2/58 (Top 4%) | ML Olympiad - Water Quality Prediction | Chris X | 2023/03/11 | v0.6.2, Tabular | |
| Rank 6/734 (Top 1%) | Tabular Regression with a Gemstone Price Dataset | Kirderf | 2023/03/06 | v0.6.2, Tabular | Kaggle Playground Series S3E8 |
| Rank 9/703 (Top 1.3%) | Tabular Regression with a Paris Housing Price Dataset | Brendan Moore | 2023/02/20 | v0.6.2, Tabular | Kaggle Playground Series S3E6 |
| 🥇 Rank 1/689 | Tabular Regression with the California Housing Dataset | Kirderf | 2023/01/09 | v0.6.1, Tabular | Kaggle Playground Series S3E1 |
To view a list of all AutoGluon research papers, please refer to our citation guide.
AMLB: An AutoML Benchmark (JMLR 2024)
- For a thorough comparison of AutoGluon and other modern AutoML systems, please refer to the 2024 JMLR paper "AMLB: An AutoML Benchmark" and the 2022 edition where AutoGluon is shown to be the state-of-the-art among AutoML systems on tabular data.
- We encourage all users to benchmark AutoGluon & other AutoML frameworks on AMLB.
- This is our preferred benchmark as it is widely accepted and trusted within the AutoML community.
The AutoML Benchmark 2025, an independent large-scale evaluation of tabular AutoML frameworks, showcases AutoGluon 1.2 as the state of the art AutoML framework. Highlights include:
- AutoGluon's rank statistically significantly outperforms all AutoML systems via the Nemenyi post-hoc test across all time constraints.
- AutoGluon with a 5 minute training budget outperforms all other AutoML systems with a 1 hour training budget.
- AutoGluon is pareto efficient in quality and speed across all evaluated presets and time constraints.
- AutoGluon with
presets="high", infer_limit=0.0001(HQIL in the figures) achieves >10,000 samples/second inference throughput while outperforming all methods. - AutoGluon is the most stable AutoML system. For "best" and "high" presets, AutoGluon has 0 failures on all time budgets >5 minutes.
Below is a sampling of some interesting papers that have cited AutoGluon.
- (2023/04/28) Benchmarking Automated Machine Learning Methods for Price Forecasting Applications
- This paper compares various traditional and AutoML methods for price forecasting problems, with AutoGluon achieving the strongest results.