DocumentationGrafana CloudAlerts and IRMMachine LearningConfigure

Configure Grafana Machine Learning

Explore ways to configure Machine Learning in Grafana Cloud to enhance your predictive analytics capabilities.

Key configuration options

Model configuration

• Grafana ML learns patterns in your data to predict time series accurately.
• Explore various model configuration options, including tuning parameters for better performance.

Source query definition

• Understand how to define the source query, representing the time series data to be modeled.
• Learn about the nuances of selecting and crafting effective source queries for accurate model training.

Setting confidence bounds

• Learn how to set confidence bounds for predicted values, enhancing the reliability of your analytics.
• Explore how to interpret and utilize confidence bounds in your monitoring.

Configure settings

The flexible Grafana Cloud environment allows you to fine-tune the way that your team leverages machine learning. Key settings to customize include:

• Model parameters: Adjust parameters to fine-tune machine learning models for optimal performance.
• Prediction thresholds: Define thresholds for predictions to align with the sensitivity required for monitoring and alerting.
• Training windows: Configure the duration and frequency of model training to adapt to evolving patterns in your data.

Explore configuration options for Machine Learning in Grafana Cloud:

• Model configuration
• LLM plugin
• Sift configuration
• Grafana Machine Learning Outlier Detection