hendrycks/outlier-exposure

Deep Anomaly Detection with Outlier Exposure (ICLR 2019)

49
/ 100
Emerging

This helps deep learning models better identify unusual or unexpected data points, often called anomalies. You provide your existing dataset along with an additional dataset of known 'outlier' examples, and the system improves your model's ability to spot anomalies it hasn't seen before. This is for machine learning engineers or researchers working with deep learning models who need to improve the reliability of anomaly detection.

575 stars. No commits in the last 6 months.

Use this if you need to significantly boost the accuracy of your deep learning model in recognizing rare or anomalous events without needing to tune for each new type of anomaly.

Not ideal if you don't have access to a large, diverse dataset of out-of-distribution examples to train with or if you are not using deep learning models.

anomaly-detection deep-learning outlier-detection model-robustness machine-learning-engineering
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 23 / 25

How are scores calculated?

Stars

575

Forks

107

Language

Python

License

Apache-2.0

Last pushed

Oct 09, 2021

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/hendrycks/outlier-exposure"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.