Expand description
ML-readiness evaluation module.
Validates that generated data is suitable for machine learning tasks including feature distributions, label quality, and graph structure.
Also provides baseline task definitions for benchmarking synthetic data.
Structs§
- Baseline
Config - Configuration for baseline evaluation.
- Baseline
Evaluation - Collection of baseline results for all tasks.
- Baseline
Result - Baseline task results.
- Baseline
Summary - Summary of baseline evaluation.
- Baseline
Task - ML baseline task definition.
- Classification
Metrics - Classification metrics for binary/multiclass tasks.
- Expected
Metrics - Expected performance metrics for a task.
- Feature
Analysis - Results of feature analysis.
- Feature
Analyzer - Analyzer for feature distributions.
- Feature
Stats - Statistics for a single feature.
- Graph
Analysis - Results of graph analysis.
- Graph
Analyzer - Analyzer for graph structure.
- Graph
Metrics - Basic graph metrics.
- Label
Analysis - Results of label analysis.
- Label
Analyzer - Analyzer for label quality.
- Label
Distribution - Distribution for a single label class.
- MLReadiness
Evaluation - Combined ML-readiness evaluation results.
- Ranking
Metrics - Ranking metrics for link prediction and recommendation.
- Regression
Metrics - Regression metrics for continuous prediction tasks.
- Split
Analysis - Results of split analysis.
- Split
Analyzer - Analyzer for train/test splits.
- Split
Metrics - Metrics for a single split.
Enums§
- Baseline
Algorithm - Baseline algorithm for a task.
- MLTask
Type - ML task type for benchmarking.
- Performance
Grade - Performance grade for baseline results.
Functions§
- get_
accounting_ baseline_ tasks - Get predefined baseline tasks for synthetic accounting data.