Numenta Logo

Anomaly Detection: Numenta Anomaly Benchmark

Numenta Anomaly Benchmark (NAB) for anomaly detection in streaming data youtube video screenshot
Video: Numenta Anomaly Benchmark (NAB) for anomaly detection in streaming data (02:23)

The First Benchmark For Evaluating Anomaly Detection In Streaming Data

The need for anomaly detection has grown, as the Internet of Things has produced a world that’s overflowing with streaming data. As these data sources continue to grow, so does the need for anomaly detection.

Uncovering anomalies allows you to:

  • Detect potential machine failures
  • Recognize changes in Twitter activity
  • Identify unexpected traffic patterns

There are different methods of anomaly detection in streaming data, but how do you measure their effectiveness? NAB is the first benchmark designed for time-series data that gives credit to finding anomalies earlier and adjusting to changed patterns.

Features

Numenta Anomaly Detection Benchmark – Heartbeat monitor

Real-World Dataset

NAB contains a dataset with real-world, labeled data files across multiple domains. We’ve accumulated this valuable data from years of working with customers to address their anomaly problems.

Numenta Anomaly Detection Benchmark – Mathematical Equations

Scoring Mechanism

We have developed a unique scoring function that rewards early detection, penalizes late or false results, and gives credit for on-line learning.

Numenta Anomaly Detection Benchmark – Geometric and Math Blueprints

Open Source Code Library

NAB is a modular, open source code base. Numenta will be working to build a community around NAB to add data files and test additional algorithms.

Resources

Numenta Anomaly Detection Benchmark – Anomaly Detection Image

Research Paper: Unsupervised Real-Time Anomaly Detection for Streaming Data

This paper introduces an anomaly detection technique using HTM and the Numenta Anomaly Benchmark (NAB). The paper also contains an analysis of the performance of ten algorithms (including HTM) on NAB.

Read Research Paper
Evaluating Real-Time Anomaly Detection youtube video screenshot
Video: Evaluating Real-Time Anomaly Detection (19:23)

Evaluating Real-Time Anomaly Detection: The Anomaly Detection: Numenta Anomaly Benchmark

Subutai Ahmad, VP Research presenting NAB and discussing the need for evaluating real-time anomaly detection algorithms. This presentation was delivered at MLConf (Machine Learning Conference) in San Francisco 2015.

See Slides
Numenta Anomaly Detection Benchmark – NAB White Paper Chart Figure

White Paper: The Anomaly Detection: Numenta Anomaly Benchmark

Why did we create this benchmark? Why is anomaly detection so hard in streaming data? This paper answers those questions and highlights how business managers can use NAB to ensure they’re getting valuable insights as early as possible.

Read Whitepaper
Numenta Anomaly Detection Benchmark – NAB Scoreboard

Research Paper: Evaluating Real-time Anomaly Detection Algorithms – the Anomaly Detection: Numenta Anomaly Benchmark

This peer-reviewed paper was accepted to the IEEE Conference on Machine Learning and Applications December 9-11, 2015 in Miami. It contains technical details on NAB, including the mathematical explanation of the scoring system.

Read Research Paper
Numenta Anomaly Detection Benchmark – NAB Github Repository

NAB Repository

This open source library contains all data files, algorithms and documentation. Use this repository to try NAB for yourself. Test your own techniques against the published algorithms and share your results.

Visit Repository
Numenta Anomaly Detection Benchmark – NAB Datasheet

Data Sheet: Anomaly Detection: Numenta Anomaly Benchmark

Download this two-page data sheet to learn more about the key components of NAB.

Download Datasheet

Try NAB for Yourself

Numenta Anomaly Detection Benchmark – NAB Chart Detail

Try the Anomaly Detection: Numenta Anomaly Benchmark

We’ve made it easy for you to try NAB. Visit the repository to test your own techniques and share your results. Use NAB to select the best algorithm for your specific application.

Contribute Your Data

We are committed to adding more real-world data files to our benchmark dataset. Do you have streaming data files with known anomalies? Contact us at nab@numenta.org to see if we can incorporate your data into a future version of NAB.