LORE_sa
  • Get started
    • Welcome to LORE_sa
    • What is LORE?
    • Installation
      • Prerequisites
      • Using virtualenv
      • Using conda
    • Quick Start
      • Basic Usage
    • Key Components
    • Choosing an Explainer
      • TabularRandomGeneratorLore
      • TabularGeneticGeneratorLore
      • TabularRandGenGeneratorLore
    • Understanding the Explanation
      • Rule
      • Counterfactuals
      • Deltas
      • Feature Importances
      • Fidelity
    • Next Steps
    • Common Pitfalls
    • Getting Help
  • Architecture and Methodology
    • Overview
    • Theoretical Foundation
    • The Four-Stage Process
      • Stage 1: Instance Encoding
      • Stage 2: Neighborhood Generation
      • Stage 3: Surrogate Training
      • Stage 4: Rule Extraction
    • Architecture Components
      • Black Box Wrapper (AbstractBBox)
      • Dataset (TabularDataset)
      • Encoder/Decoder (EncDec)
      • Neighborhood Generator (NeighborhoodGenerator)
      • Surrogate Model (Surrogate)
      • Rule and Expression Classes
    • Workflow Diagram
    • Best Practices
      • Choosing num_instances
      • Interpreting Fidelity
      • Handling Low Fidelity
      • Feature Importance Interpretation
    • Computational Complexity
      • Time Complexity
      • Space Complexity
      • Typical Running Times
    • References
      • Primary Paper
      • Related Work
      • Key Differences from LIME
    • Implementation Notes
      • Thread Safety
      • Memory Considerations
      • Reproducibility
  • Tabular explanation example
    • Learning and explaining German Credit Dataset
    • Loading and preparation of data
      • Learning a Random Forest classfier
    • Explaining the prediction
      • SHAP explainer
      • LORE explainer
      • LIME explainer
    • Learning a different model
      • Learning a Logistic Regressor
    • Explaining the prediction
      • LORE explainer
      • LIME explainer
  • Modules
    • lore_sa.bbox: BlackBox abstract class
      • lore_sa.bbox.AbstractBBox
        • AbstractBBox
    • lore_sa.dataset: Dataset class
      • lore_sa.dataset.Dataset
        • Dataset
      • lore_sa.dataset.TabularDataset
        • TabularDataset
      • lore_sa.dataset.utils
        • prepare_bank_dataset()
    • lore_sa.neighgen: Neighborhood Generator classes
      • lore_sa.neighgen.RandomGenerator
        • RandomGenerator
      • lore_sa.neighgen.GeneticGenerator
        • GeneticGenerator
    • lore_sa.discretizer: Discretizer classes and functions
      • lore_sa.discretizer.Discretizer
        • Discretizer
      • lore_sa.discretizer.RMEPDiscretizer
        • RMEPDiscretizer
    • lore_sa.encoder_decoder: Encoder/Decoder classes and functions
      • lore_sa.encoder_decoder.EncDec
        • EncDec
      • lore_sa.encoder_decoder.ColumnTransformerEnc
        • ColumnTransformerEnc
    • lore_sa.explanation: Explanation classes and functions
      • lore_sa.explanation.Explanation
        • Explanation
      • lore_sa.explanation.ExplanationEncoder
        • ExplanationEncoder
      • lore_sa.explanation.ImageExplanation
        • ImageExplanation
      • lore_sa.explanation.MultilabelExplanation
        • MultilabelExplanation
      • lore_sa.explanation.TextExplanation
        • TextExplanation
      • lore_sa.explanation.json2explanation
        • json2explanation()
    • lore_sa.rule: Rule classes and functions
      • lore_sa.rule.Rule
        • Rule
      • lore_sa.rule.RuleEncoder
        • RuleEncoder
      • lore_sa.rule.json2rule
        • json2rule()
      • lore_sa.rule.json2expression
        • json2expression()
      • lore_sa.rule.Expression
        • Expression
    • lore_sa.surrogate: Surrogate classes and functions
      • lore_sa.surrogate.Surrogate
        • Surrogate
      • lore_sa.surrogate.DecisionTreeSurrogate
        • DecisionTreeSurrogate
    • lore_sa.util: Util functions
      • lore_sa.util
        • best_fit_distribution()
        • sigmoid()
        • vector2dict()
LORE_sa
  • Search

© Copyright 2023, Kode srl.

Built with Sphinx using a theme provided by Read the Docs.