Skip to content

Predictive AI reference

The following sections provide reference content that supports working with predictive and time-aware experiments:

Topic Description
Data partitioning Describes validation types and data partitioning methods.
Feature lists Shows details of working with DataRobot-generated and custom feature lists, as well as where in the platform you can create and manage them.
Modeling algorithms Lists supervised and unsupervised modeling algorithms supported by DataRobot.
Modeling process Describes modeling modes, two-stage models, and data summary information.
Model recommendation process Describes the steps involved in DataRobot's selection of a recommended model.
Leaderboard reference Provides a reference table of the badges that display in the Leaderboard and the Blueprint repository, model icons, and other Leaderboard indicators.
Optimization metrics Briefly describes all metrics available for model building.
SHAP reference Provides details of SHapley Additive exPlanations, the coalitional game theory framework.
Feature Associations Explains about associations, understanding the mutual information and Cramer's V metrics, and how associations are calculated.
Insurance-specific settings Describes Exposure, Count of events, and Offset options, configured in advanced settings.
Sliced insights Describes sliced insights where you can view and compare insights based on segments of a project’s data.
Bias and Fairness reference Provides an overview of the methods used to calculate fairness and to identify biases in the model's predictive behavior.
Time series reference Provides reference material explaining the DataRobot framework for implementing time series modeling and see a variety of deep-dive reference material for DataRobot time series modeling.
Eureqa advanced tuning Describes how to modify building blocks, customize the target expression, and modify other model parameters for Eureqa models.
Composable ML reference Provides information on blueprints in the AI Catalog, model metadata, feature considerations, and a sentiment analysis example.
Visual AI reference Provides workflow and reference materials for including images as part of your DataRobot experiments.
Export charts and data Explains about downloading created insights.
Worker queue (NextGen) Helps to understand modeling workers and how to troubleshoot issues in NextGen.
Worker queue (Classic) Helps to understand modeling workers and how to troubleshoot issues in Classic.
XEMP qualitative strength Describes the calculations used to determine XEMP qualitative strength.
AI Report (Classic only) Describes how to create a report of modeling results and insights.