# numenta/NAB

The Numenta Anomaly Benchmark

Repository: https://github.com/numenta/NAB
Canonical: https://ross.abutalabs.com/products/nab
Language: Jupyter Notebook
License: MIT
License Family: permissive
Last push: 2024-12-03T17:49:50+00:00

## Health v2 (maintenance only)
Score: 23/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 4502, "days_push": 638, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2103, forks 873 (observed 2026-08-28T04:06:13.781936+00:00)

## What it is
The Numenta Anomaly Benchmark (NAB) is a benchmark for evaluating anomaly detection algorithms on streaming, real-time time-series data. It includes over 50 labeled real-world and artificial time-series data files, scoring scripts, and a real-time-aware scoring mechanism.

## Use cases
- benchmark my anomaly detection algorithm on streaming time series
- find labeled time-series data with anomalies for testing
- compare anomaly detectors with a real-time scoring metric
- evaluate how quickly a detector catches anomalies with few false positives
- run a standard anomaly detection benchmark in python

## When to choose
- you need a standardized, labeled benchmark for streaming anomaly detection
- you want comparable scores against published detectors like HTM or Skyline
- you need real-world time-series anomaly data for research

## When to avoid
- you need image, text, or log anomaly detection benchmarks
- you want a production anomaly detection service rather than a benchmark
- you need actively developed tooling with frequent updates

## Facets
- artifact type: dataset
- maturity: maintenance
- function: benchmarking, machine-learning, data-science
- domain: machine-learning, data-science, analytics, time-series
- platform: python, cross-platform
- tags: anomaly-detection, time-series, streaming-data, benchmark, labeled-data

## Member repositories
- numenta/NAB (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:13.781936+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T02:54:15.187379+00:00, confidence not recorded.
  - readme: https://github.com/numenta/NAB (fetched 2026-08-28T04:06:13.781936+00:00, sha f9a24fdecd4b)
- Data as of 2026-08-30T08:39:29.467469+00:00.