AlphaX
  • WORKSPACE
    • Data Table List View
      • AI View
    • Map
  • Analysis
    • Overview
    • Anomalies
      • Batch Detection
      • Change Point Detection
      • Streaming Detection
      • User interface
        • Anomaly List View
        • Anomaly Edit View
        • Anomaly Chart View
        • Anomaly Share View
      • Analysis List View
    • TOOLBOX
      • Analysis
        • Charts
          • Charts View
          • Drawings View
          • Data View
        • Table
        • Anomaly
      • Reports
        • Reports List View
          • Create a New Report
        • Reports Manage View
      • Device
        • Devices List View
          • Create New Device Registration
        • Devices Manage View
      • API Endpoints
        • API Endpoints List View
          • Create a New Feed
        • API Endpoints Manage View
      • Integrations
        • Integrations List View
          • App
        • Integrations Manage View
      • Dashboards
        • Dashboards List View
          • Create a New Dashboard
        • Dashboards Manage View
      • Alerts
        • Events Hub List View
        • Events Hub Manage View
        • Event Data Format
        • Event Acknowledgement
      • Calculations
        • Calculations List View
          • Create a New Calculation
        • Calculations Manage View
      • Controls
        • Controls List View
          • Create a New Control Element
        • Controls Manage View
  • Data Hub
    • API Queries
      • API Queries List View
      • API Queries Modal View
      • Testing API Queries
      • Common Issues & Troubleshooting
    • ADMINISTRATION
      • People
        • Users
          • Users List View
            • Create a New User Account
          • Users Manage View
        • Roles
          • Roles List View
            • Create a New Access Role
          • Roles Manage View
        • Contacts
          • Contacts List View
            • Create New Contact
          • Contacts Manage View
      • Configuration
        • Asset Hierarchy
          • Asset Hierarchy List View
            • Create New Entity
          • Asset Hierarchy Manage View
        • Attributes
          • Asset Classes
            • Asset Classes List View
              • Create New Asset Class
            • Asset Classes Manage View
          • Battery Types
            • Battery Types List View
              • Create New Battery Type
            • Battery Types Manage View
          • Device Categories
            • Device Categories List View
              • Create New Device Category
            • Device Categories Manage View
          • Feed Types
            • Feed Types List View
              • Create a New Feed Type
            • Feed Types Manage View
          • Manufactures
            • Manufacturers List View
              • Create a New Manufacture
            • Manufactures Manage View
          • Models
            • Models List View
              • Create a New Model
            • Models Manage View
          • Network Types
            • Network Types List View
              • Create New Network Type
            • Network Types Manage View
          • Power Sources
            • Power Sources List View
              • Create A New Power Source
            • Power Sources Manage View
          • Sensor Types
            • Sensor Types List View
              • Create New Sensor Type
            • Sensor Types Manage View
          • Status Labels
            • Status Labels List View
              • Create New Status Label
            • Status Labels Manage View
          • Suppliers
            • Suppliers List View
              • Create New Supplier
            • Suppliers Manage View
      • System Logs
        • System Logs List View
        • Email Message Log
        • SMS Message Log
        • API Transaction Log
        • Create a New Service Log Report
Powered by GitBook
On this page
  • What is it for?
  • How does it work?
  1. Analysis
  2. Anomalies

Batch Detection

Batch detection analyses groups of data to identify anomalies by comparing data points against expected values, flagging outliers.

PreviousAnomaliesNextChange Point Detection

Last updated 8 months ago

What is it for?

Batch anomaly detection is used to identify anomalies in data that is processed in groups or batches rather than in real-time. This method is applied to a collected set of data, typically over a specific period, to analyze and detect any outliers or unusual patterns.

In IoT/OT environments, batch anomaly detection is useful for analyzing historical data to identify trends, seasonal variations, or anomalies that may not be immediately apparent in real-time. It's often used for periodic reporting, trend analysis, and retrospective investigations of system behavior.

Output of Batch Detection Anomoly

How does it work?

Identifies anomalies across the entire time series data.

Sends the most recent 1000 data points to the AlphaX intelligence engine to build a model, then analyses each data point based on that model.

Parameters

Feature
Description

Sensitivity

A numerical value (0% - 99%) that adjusts the tolerance of Anomaly Detection. Lowering the sensitivity leads to fewer anomalies detected.

Max Anomaly Ratio

A numerical value (0% - 49%) that indicates the maximum ratio of anomalies to be detected from a time series. For example, if it's set to default 0.25 (25%), for a time series with 100 points, the maximum number points identified as an anomaly would be 25.

Output

Feature
Description

Sensor

The actual measured value from the sensor or system

Anomaly

Boolean: If the data point is Identified as an anomaly

Expected

The expected or predicted values of the data points in the time series

Lower Margin

Lower margin of each date point, used to calculate Lower Boundary.

Upper Margin

Upper margin of each date point, used to calculate Upper Boundary.

Boundaries

Boundaries in batch mode anomaly detection define the expected limits for data values within a batch. These boundaries help the user to visualise whether a data point is normal or anomalous but are not absolute.

High (Lower Boundary): Calculated as the expected value plus the average of the current and previous 2 upper margins.

Low (Upper Boundary): Calculated as the expected value minus the average of the current and previous 2 lower margins.