Skip to main content

> ML_LITERATURE // LUNDBERG-2017-UNIFIED-APPROACH-INTERPRETING-MODEL-PREDICTIONS-SHAP_v1.0

A Unified Approach to Interpreting Model Predictions (SHAP)

Scott M. Lundberg, Su-In Lee · Advances in Neural Information Processing Systems (NeurIPS) (2017)

foundational2017industry-standardthirdPartyReproduced

Principal Contribution

Unified LIME, DeepLIFT, and Layer-wise Relevance Propagation under game-theoretic Shapley values (SHAP), proving it is the unique attribution method satisfying local accuracy and consistency.

Operational Relevance

Serves as qualified reference for implementing task-model-explainability in production systems.

Assumptions

  • Underlying computational topology and mathematical bounds adhere to established convexity/smoothness guarantees

Limitations

  • Hardware runtime speedups, privacy budgets, and convergence depend on hyperparameters and network communication limits

Connected Algorithms, Architectures & Tools

Related Algorithms:
Related Architectures:
Implementing Libraries: