Luminaire is a python package that provides ML driven solutions for monitoring time series data
Project description
Luminaire: A hands-off Anomaly Detection Library
Table of contents
What is Luminaire
Luminaire is a python package that provides ML driven solutions for monitoring time series data. Luminaire provides several anomaly detection and forecasting capabilities that incorporate correlational and seasonal patterns in the data over time as well as uncontrollable variations.
Please see the full Luminaire documentation for detailed descriptions of the methods.
Quick Start
Luminaire can be installed from PyPI.
pip install luminaire
Time Series Outlier Detection Workflow
Luminaire outlier detection workflow can be divided into 3 major components.
Data Preprocessing and Profiling Component
This component can be called for preparing a time series before training an anomaly detection model. This step applies all necessary fixes (missing data imputation, identifying and removing recent outliers from training data, necessary mathematical transformations, data truncation based on recent change points etc) and also generates profiling information (hostorical change points, trend changes etc) for the training data.
The profiling information for time series data generates important information for an offline time series data and can be used to monitor irregular longer term swings or data drifts.
Modeling Component
This components performs time series model training based on the user specified configuration OR optimized configuration (see Luminaire hyperparameter optimization). Luminaire model training is integrated with different structural time series models as well as filtering based models. See Luminaire outlier detection for more information.
Luminaire modeling step can be called after the data preprocessing and profiling step to perform necessary data preparation before training.
Configuration Optimization Component
Luminaire is integrated with configuration optimization capability for the hands-off anomaly detection approach where the user needs to provide almost no configuration for any type of time series data. This step can be combined with the preprocessing and modeling for any auto configured anomaly detection use case. See fully automatic outlier detection for a detailed walkthrough.
Anomaly Detection for High Frequency Time Series
Luminaire can monitor a set of data points over windows instead of tracking individual. This approach becomes relevant for streaming use cases where individual fluctuations is not of a concern but the data is monitored for more sustained fluctuations. See anomaly detection for streaming data for detailed information.
Contributing
Want to help improve Luminaire? Check out our contributing documentation.
Acknowledgements
This project has leveraged methods described in the following scientific publications:
- Soule, Augustin, Kavé Salamatian, and Nina Taft. "Combining filtering and statistical methods for anomaly detection." Proceedings of the 5th ACM SIGCOMM conference on Internet Measurement. 2005.
Development Team
Luminaire is developped and maintained by Sayan Chakraborty, Smit Shah, Kiumars Soltani, Luyao Yang, Anna Swigart, Kyle Buckingham and many other contributors from the Zillow Group A.I. team.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for luminaire-0.1.0.dev6-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 983d29bc864f36e93cb7234e913a6d50b2a66a59021ff02c32f03911cacccb9d |
|
MD5 | 226208235b4c5c7cd50ca6c0948924a0 |
|
BLAKE2b-256 | 5bb78ea12d90337695286eea0770ad2161db3629245315f9433d27669382b2b8 |