Predictive Models

H2O Eval Studio provides robust interpretability of predictive machine learning models.

Models

• Interpreting Models
  • Explainable Model
  • Interpreting Models using the Command Line Interface
  • Interpreting Models using the Python API

Datasets

• Interpreting Datasets
  • Explainable Dataset

Interpretation

• Interpretation Parameters

Explainers

• Explainers
  • Explainer Parameters
  • Surrogate Decision Tree Explainer
  • Residual Surrogate Decision Tree Explainer
  • Shapley Summary Plot for Original Features (Naive Shapley Method) Explainer
  • Shapley Values for Original Features (Kernel SHAP Method) Explainer
  • Shapley Values for Original Features of MOJO Models (Naive Method) Explainer
  • Shapley Values for Transformed Features of MOJO Models Explainer
  • Partial Dependence/Individual Conditional Expectations (PD/ICE) Explainer
  • Partial Dependence for 2 Features Explainer
  • Residual Partial Dependence/Individual Conditional Expectations (PD/ICE) Explainer
  • Disparate Impact Analysis (DIA) Explainer
  • Morris sensitivity analysis Explainer
  • Friedman’s H-statistic Explainer
  • Dataset and Model Insights Explainer
  • Adversarial Similarity Explainer
  • Backtesting Explainer
  • Calibration Score Explainer
  • Drift Detection Explainer
  • Segment Performance Explainer
  • Size Dependency Explainer