TreeSHAP
PulseAugur coverage of TreeSHAP — every cluster mentioning TreeSHAP across labs, papers, and developer communities, ranked by signal.
4 day(s) with sentiment data
-
New framework detects AI model drift missed by local XAI methods
Researchers have developed a new Framework for Model Monitoring and Observability (FMMO) to address the critical risk of post-deployment drift in AI models. Traditional Explainable AI (XAI) methods, such as TreeSHAP, ca…
-
GraphFAS system automates graph feature selection for fraud detection
A new system called GraphFAS has been developed for automated graph feature generation and selection, specifically designed for industrial transaction networks. This system addresses the limitations of traditional exper…
-
New AI system explains its detection of machine-generated text
Researchers have developed NOTAI.AI, a novel system for detecting machine-generated text that goes beyond simple binary classification. This system provides explainability by highlighting the specific features and signa…
-
Explainable AI identifies broadband adoption disparities across US tracts
Researchers have developed an explainable machine learning framework to identify disparities in broadband adoption across the United States. The model, trained on socioeconomic and demographic data, achieved strong pred…
-
New Causal Attribution Score (CAS) enhances AI explainability
Researchers have introduced the Causal Attribution Score (CAS), a novel framework for causal explanation in artificial intelligence. CAS distinguishes itself by attributing intervention effects on real-world outcomes, r…
-
IoT Intrusion Detection: Beyond Accuracy to Explanation Cost and Stability
A new study published on arXiv evaluates machine learning models for Internet of Things (IoT) intrusion detection, focusing beyond just accuracy to include explanation cost, stability, and utility. Researchers construct…
-
New fraud detection system combines graph features, LLM agents
Researchers have developed a layered approach for fraud detection that combines gradient-boosted classifiers with graph-derived structural features and an LLM investigation agent. This system was tested on the PaySim da…
-
XGBoost model deciphers stock market predictability with behavioral signals
Researchers have developed an interpretable machine learning pipeline to analyze stock return predictability in large-scale financial markets, specifically focusing on China's A-share market. Using an XGBoost model with…
-
AI pipeline enhances financial fraud detection with LLM-generated narratives
Researchers have developed an end-to-end pipeline for identifying money mule accounts, a critical component in combating financial fraud. The system utilizes a LightGBM classifier with 280 features, a TreeSHAP layer for…
-
SHAP-weighted fusion method shows promise for emotion and sentiment recognition
Researchers have analyzed the effectiveness of SHAP-weighted cross-modal expert fusion ("xgaf") for emotion and sentiment recognition. The study found that using sum-abs reduction for SHAP attribution magnitudes, partic…
-
New metric measures how AI security classifier explanations degrade under attack
A new research paper introduces the Explainability Stability Index (ESI) to measure how adversarial attacks affect the explanations of cybersecurity classifiers. The study, which extends prior work to Random Forest and …
-
AI model deciphers stock market predictability using behavioral signals
Researchers have developed an interpretable machine learning pipeline to break down stock market predictability into factor contributions. Applying an XGBoost model with TreeSHAP attribution to Chinese A-share stocks fr…
-
New LMDI+ method enhances interpretability for tree-based models
Researchers have developed Local MDI+ (LMDI+), a new method for quantifying feature importances in tree-based models for individual samples. Unlike existing approximation-based methods, LMDI+ leverages the internal stru…
-
Quadrature-TreeSHAP offers faster, more stable AI model explanations
Researchers have developed Quadrature-TreeSHAP, a novel method for explaining tree ensemble predictions that is depth-independent and more numerically stable than existing approaches. This new technique extends naturall…