• Help
  • 日本語
  • Home
  • Predict ML
  • Getting Started
    • About Predict
    • Using Predict
    • Best practices
    • Prerequisites
    • Terminology
  • Define a strategy
    • Decide what to predict
    • Attribute readiness
    • Prepare your data
    • Start training
  • Create training models
    • Add a model
    • Review your model
  • Evaluate and retrain models
    • Evaluate trained models
    • Strength scores
    • The Confusion Matrix
    • The ROC/AUC curve
    • Probability distribution
    • View model status
    • Retrain a model
  • Deploy models
    • Deploy a model
    • Undeploy a model
    • Deployed model health
    • Model retraining recommendations
    • Delete a model
  • Create audiences
    • Audience considerations
    • Create audiences with Predict
  • Advanced topics
    • Machine learning vs. artificial intelligence
    • Machine learning concepts and technology
  • Home > 
  • Predict ML

Evaluate and retrain models

Evaluate trained models

This article provides an overview of how to evaluate your trained version before deploying your model. Use the following sections as a guide to view model status and the potential strength of predictions for your model before you deploy.

Strength scores

This article provides detailed information about model scoring techniques and formulas used to assign scores and ratings to trained and deployed models in Tealium Predict ML.

The Confusion Matrix

This article describes the Confusion Matrix and how to use it to evaluate a trained model.

The ROC/AUC curve

This article describes the ROC/AUC curve and how to use it as a performance measurement of a trained model.

Probability distribution

This article describes how to use the probability distribution to interpret trained models.

View model status

This article describes how to view your model status.

Retrain a model

This article describes how to retrain a model after evaluating the model and determining changes that are required to improve predictions.

  • tealium_footer_logo_01
  • Support
  • Tealium University
  • Knowledge Base
  • Support Desk
  • Company
  • Tealium.com
  • Blog
  • Careers
  • Home
  • Guides
  • iQ Tag Management
  • EventStream API Hub
  • AudienceStream CDP
  • Moments API
  • Functions
  • Data Connect
  • Predict ML
  • DataAccess
  • Tealium Insights
  • Client-Side Tags
  • Server-Side Connectors
  • Platforms
  • API
  • Administration
  • Tealium Tools
  • Tealium for Pharma
  • Tealium Identity Partners
  • Technology Partners
  • Partner Integrations
  • Glossary
  • Early Access
  • Release Notes
  • About Predict
  • Using Predict
  • Best practices
  • Prerequisites
  • Terminology
  • Decide what to predict
  • Attribute readiness
  • Prepare your data
  • Start training
  • Add a model
  • Review your model
  • Evaluate trained models
  • Strength scores
  • The Confusion Matrix
  • The ROC/AUC curve
  • Probability distribution
  • View model status
  • Retrain a model
  • Deploy a model
  • Undeploy a model
  • Deployed model health
  • Model retraining recommendations
  • Delete a model
  • Audience considerations
  • Create audiences with Predict
  • Machine learning vs. artificial intelligence
  • Machine learning concepts and technology