Quantifying Sensitivity for Tree Ensembles: A symbolic and compositional approach

2026-05-13 · Source: Takara TLDR - Daily AI Papers · Field: Technology & Digital — Artificial Intelligence & Machine Learning · Depth: Expert, quick

Summary

A new algorithmic technique, XCount, has been developed to quantify the sensitivity of Decision Tree Ensembles (DTEs), which are widely used in AI classification tasks, including safety-critical domains. The method addresses the problem of determining if small changes in input features can cause misclassification. XCount discretizes the DTE input space and enumerates sensitive regions, providing a quantitative measure of sensitivity with certified error and confidence bounds. The approach encodes the problem as an Algebraic Decision Diagram (ADD) and decomposes it into subproblems for efficient, compositional, and scalable computation. Experimental evaluations on benchmarks of varying tree numbers and depths demonstrate that XCount achieves significant speedup compared to model counters and scales effectively with increasing ensemble sizes.

Key takeaway

For AI Scientists and Research Scientists working with Decision Tree Ensembles in safety-critical applications, understanding model sensitivity is crucial. XCount offers a robust, efficient method to quantify this sensitivity, providing certified error and confidence bounds. You should consider integrating XCount into your model verification pipeline to identify and mitigate potential misclassification risks arising from minor feature perturbations, especially for large ensembles.

Key insights

XCount quantifies Decision Tree Ensemble sensitivity by discretizing input space and enumerating susceptible regions efficiently.

Principles

• Discretize input space for sensitivity analysis.
• Encode DTE sensitivity as an Algebraic Decision Diagram (ADD).

Method

The method encodes the sensitivity problem as an Algebraic Decision Diagram (ADD), then splits it into subproblems for compositional and scalable computation, enumerating susceptible regions with certified error and confidence bounds.

In practice

• Use XCount for DTE sensitivity analysis.
• Apply ADD-based encoding for similar verification tasks.

Topics

• Decision Tree Ensembles
• Model Sensitivity
• AI Classification
• Algebraic Decision Diagrams
• XCount Tool

Best for: AI Scientist, Research Scientist

Related on AIssential

• Quantifying Sensitivity for Tree Ensembles: A symbolic and compositional approach — A new algorithm efficiently quantifies DTE sensitivity by discretizing input space and using Algebraic Decision Diagrams.
• Quantifying Sensitivity for Tree Ensembles: A symbolic and compositional approach — The paper introduces XCount, a compositional, ADD-based algorithm for quantifying DTE sensitivity with certified approximate counts, outperforming prior methods.
• CauTion: Knowing When to Trust LLMs for Ensemble Causal Discovery — CauTion integrates LLM domain knowledge with statistical ensembles using trust calibration to improve causal discovery accuracy.
• Ribbon: Scalable Approximation and Robust Uncertainty Quantification — Ribbon offers scalable, robust uncertainty quantification by linearizing bootstrap reweighting around a single model fit, avoiding costly retraining.
• Optimal Deterministic Multicalibration and Omniprediction — Deterministic multicalibration can achieve minimax-optimal sample complexity, resolving a key open problem in trustworthy machine learning.
• [D] Risk of using XGB models — Weak features in ensemble models can lead to overfitting and reduced robustness, despite aggregation.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Takara TLDR - Daily AI Papers.