Skip to main content
Documentation
Documentation
Administrator Guide
Last Updated: 2025-02-07
Steps: Set Up Machine Learning for Time Anomalies

Steps: Set Up Machine Learning for Time Anomalies

  • Deploy Workday Time Tracking. We recommend you deploy Time Tracking at least 1 year before implementing time anomaly detection.
  • Create time approval templates.
You can enable Workday to use machine learning to detect atypical time entries so that you can:
  • Create custom reports to view time anomaly scores and top time anomaly reasons for time blocks.
  • Access the
    Time Anomaly Trends
     report to view time anomaly trends by supervisory organization.
  • Enable managers and timekeepers to see and take action on time anomalies within the
    Review Time
     report.
  1. Enable Innovation Services Feature and Machine Learning Data Contributions.
    On the
    Innovation Services and Data Selection Opt-In
     report, select the
    HCM: Workforce Management Machine Learning Services
     section.
    To build a machine learning model that can detect time anomalies, you must opt in to all of these categories:
    • Time Configuration Data
    • Worker Data
    • Historical Schedule Block Data
    • Worker Time and Absence Data
  2. Access the
    Edit Tenant Setup - Machine Learning
     task.
    Select the region in which Workday hosts data used for improvement and personalization of machine learning and analytics functionality..
    Security:
    Set Up: Tenant Setup - Machine Learning
     in the System functional area.
  3. (Optional) Access the
    Edit Time Approval Template
     task.
    Select the
    Include Time Anomalies
     check box to display time anomalies in the
    Review Time
     report.
    Security:
    Set Up: Time Tracking
     in the Time Tracking functional area.
We recommend that you train the machine learning model in the Production tenant for at least 3 weeks. When you train in a Production tenant, you ensure that the model has a constant stream of real data to train on so that it can learn patterns of acceptable and unacceptable time blocks.
If you need to test in a nonproduction tenant, you can enable time anomaly detection in Implementation tenants. If you began testing in an Implementation tenant before 2024-10-04, you now need to enable on-demand data extraction to continue regular data extraction.