Microsoft Anomaly Detector Bot

Boost your business systems’ reliability with early problem detection using artificial intelligence

Top Benefits

  • Speed your time to insights with simple setup and real-time anomaly detection.
  • Save time and stay focused on fixing problems as soon as they surface with prebuilt API integration
  • Maximize accuracy with Microsoft’s powerful inference engine

Tasks

  • Reads time-series data set and automatically selects the right algorithm to maximize accuracy.
  • Automatic detection eliminates the need for labeled training data.
  • Performs both real time and batch anomaly detection

Microsoft Anomaly Detector Bot looks at your time-series data set and automatically selects the best algorithm from the model gallery to ensure high accuracy for your specific scenario: business incidents, monitoring IoT device traffic, managing fraud, responding to changing markets, and more.

Use this bot in your automation for business incidents, monitoring IoT device traffic, managing fraud, responding to changing markets and more.

Access Now

Free

Bot Security Program
Level 1
Applications
Business Process
Category
Downloads
35
Vendor
Automation Type
Bot
Last Updated
May 19, 2021
First Published
March 1, 2020
Platform
Automation 360
Support

See the Bot in Action

Overview of bot
Bot in action
Run successful
PREV NEXT
Overview of bot
Bot in action
Run successful

Setup Process

Install

Download the Bot and follow the instructions to install it in your AAE Control Room.

Configure

Open the Bot to configure your username and other settings the Bot will need (see the Installation Guide or ReadMe for details.)

Run

That's it - now the Bot is ready to get going!

Requirements and Inputs

  • Endpoint URL: Get the endpoint when we created an Anomaly Detector resource on Azure portal
  • Subscription key: Get any subscription between key1 or key2 when we created an Anomaly Detector resource on Azure portal
  • Input File: time-series data eg: “C:Desktopdatarequest-data.csv”, : The input file is a csv file with one
  • X: TimeStamp (should be in ISO9601 format) eg: “2018-03-01T00:00:00Z”
  • Y: Number (Should be in Float or Integer) eg: 32858923
  • Max Anomaly Ratio (between 0.0-0.49): 0.4, Advance parameter to fine tune the model.The 'maxAnomalyRatio' field must be
  • Sensitivity (between 0-99): The lower sensitivity means more freedom to avoid the expected values