Skip to main content

Fast AutoML

Most autoML packages aim for exceptional performance but need to train for an exceptional amount of time. Fast-autoML aims for reasonable performance in a reasonable amount of time.

Fast-autoML includes additional utilities, such as tools for comparing model performance by repeated cross-validation.

Installation

$ pip install fast-automl

Quickstart

from fast_automl.automl import AutoClassifier

from sklearn.datasets import load_digits
from sklearn.model_selection import cross_val_score, train_test_split

X, y = load_digits(return_X_y=True)
X_train, X_test, y_train, y_test = train_test_split(X, y, shuffle=True, stratify=y)

clf = AutoClassifier(ensemble_method='stepwise', n_jobs=-1, verbose=True).fit(X_train, y_train)
print('CV score: {:.4f}'.format(cross_val_score(clf.best_estimator_, X_train, y_train).mean()))
print('Test score: {:.4f}'.format(clf.score(X_test, y_test)))

This runs for about 6-7 minutes and typically achieves a test accuracy of 96-99% and ROC AUC above .999.

from fast_automl.automl import AutoRegressor

from sklearn.datasets import load_diabetes
from sklearn.model_selection import cross_val_score, train_test_split

X, y = load_diabetes(return_X_y=True)
X_train, X_test, y_train, y_test = train_test_split(X, y, shuffle=True)

reg = AutoRegressor(n_jobs=-1, verbose=True).fit(X_train, y_train)
print('CV score: {:.4f}'.format(cross_val_score(reg.best_estimator_, X_train, y_train).mean()))
print('Test score: {:.4f}'.format(reg.score(X_test, y_test)))

This runs for about 30 seconds and typically achieves a test R-squared of .47-.53.

Citation

@software{bowen2021fast-automl,
  author = {Dillon Bowen},
  title = {Fast-AutoML},
  url = {https://dsbowen.github.io/fast-automl/},
  date = {2021-02-05},
}

License

Users must cite this package in any publications which use it.

It is licensed with the MIT License.

Acknowledgments

This package and its documentation draw heavily on scikit-learn.

Metadata

Release files for fast-automl 0.0.1

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for fast-automl 0.0.1
File Size Uploaded Provenance Actions
fast-automl-0.0.1.tar.gz 16.0 kB Without provenance Details for fast-automl-0.0.1.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for fast-automl 0.0.1
File Interpreter ABI Platform Actions
fast_automl-0.0.1-py3-none-any.whl Python 3 none any Details for fast_automl-0.0.1-py3-none-any.whl

Total release size: 36.7 kB

Release files / fast-automl-0.0.1.tar.gz

Download URL fast-automl-0.0.1.tar.gz
Size 16.0 kB
Tags Source
SHA-256 checksum
How to use checksums
9baea86ba557ff2b2f019033f3326522c1a1c88d18751454c29880b2805d6bcd
BLAKE2b-256 checksum
How to use checksums
91a510f3f20bcee93bc99b561822d9eb9070d56796473216eb9d723098cd47e5
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/3.3.0 pkginfo/1.7.0 requests/2.25.1 setuptools/49.2.1 requests-toolbelt/0.9.1 tqdm/4.56.0 CPython/3.9.1

Release files / fast_automl-0.0.1-py3-none-any.whl

Download URL fast_automl-0.0.1-py3-none-any.whl
Size 20.7 kB
Tags Python 3
SHA-256 checksum
How to use checksums
a63f9655533816b0d1afb2ca654c64b0a2a06c6514116d8f57e1ad3e48d6cec6
BLAKE2b-256 checksum
How to use checksums
841f26c8c79a102e4f65b09eddccbf2f42eb347566027fbed23fa98b3335b5b8
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/3.3.0 pkginfo/1.7.0 requests/2.25.1 setuptools/49.2.1 requests-toolbelt/0.9.1 tqdm/4.56.0 CPython/3.9.1

Security for fast-automl 0.0.1

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

No provenance found

None of the files in this release were uploaded with provenance metadata.

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Release history Release notifications | RSS feed

This release

0.0.1 This release

2 release files No provenance

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page