What types of anomalies can it identify?

The skill can detect point outliers, level shifts (sudden changes in the mean), and trend breaks in your time series data.

Does this skill require training a machine learning model?

No, it uses TimeGPT’s zero-shot capabilities, allowing you to detect anomalies immediately without any prior model training.

What data format is required for the anomaly detector?

The skill expects a CSV file with three specific columns: unique_id (identifier), ds (timestamp), and y (numeric value).

Do I need an API key to use this skill?

Yes, you must have a valid Nixtla TimeGPT API key stored in the NIXTLA_TIMEGPT_API_KEY environment variable.

Nixtla Anomaly Detector

Name: Nixtla Anomaly Detector
Author: intent-solutions-io

byintent-solutions-io

データサイエンスとML

Detects anomalies, outliers, and trend breaks in time series data using TimeGPT without the need for manual model training.

概要

The Nixtla Anomaly Detector skill empowers developers and data scientists to identify unusual patterns in time series data instantly using the power of TimeGPT. By integrating Nixtla's zero-shot forecasting capabilities, this skill automatically flags outliers, level shifts, and structural breaks in CSV-formatted datasets, providing both granular data outputs and visual plots for immediate analysis. It is an ideal solution for monitoring website traffic, sales performance, or any metrics where identifying deviations from the norm is critical for operational success.

主な機能

Detailed CSV and summary reports for seamless data integration
Support for multi-series datasets via the unique_id column schema
Automated generation of visualization plots highlighting anomalies in red
0 GitHub stars
Zero-shot anomaly detection using TimeGPT's pre-trained models
Identification of multiple anomaly types including outliers and level shifts

ユースケース

Detecting unusual patterns in server or website traffic logs
Identifying structural breaks in financial or sensor data for maintenance
Monitoring e-commerce sales for unexpected spikes or drops in demand

概要

主な機能

Detailed CSV and summary reports for seamless data integration
Support for multi-series datasets via the unique_id column schema
Automated generation of visualization plots highlighting anomalies in red
0 GitHub stars
Zero-shot anomaly detection using TimeGPT's pre-trained models
Identification of multiple anomaly types including outliers and level shifts

ユースケース

Detecting unusual patterns in server or website traffic logs
Identifying structural breaks in financial or sensor data for maintenance
Monitoring e-commerce sales for unexpected spikes or drops in demand