New method quantifies robustness of Shapley values in tree-based models

Researchers have developed a new method using an imprecise Dirichlet model (IDM) to analyze the robustness of Shapley values in decision trees and random forests. This approach quantifies and examines interval-valued Shapley values when unannotated instances are introduced into tree leaves. The method allows for the computation of these interval-valued Shapley values based on pessimistic and averaging principles, and can be extended to Banzhaf values. Experiments demonstrate the behavior of these interval-valued Shapley values and their utility in debiasing uninformative features. AI

IMPACT Enhances understanding of feature attribution robustness in common machine learning models.

RANK_REASON The cluster contains an academic paper detailing a new method for analyzing model explainability. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• arXiv
• Banzhaf values for cooperative games with fuzzy characteristic function
• decision tree
• Hugging Face
• imprecise Dirichlet model (IDM)
• interval-valued Shapley values
• random forest
• Shapley values
• tabular data
• Tree-based models for poverty estimation

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →