Researchers have developed a new Machine Learning framework using XGBoost to emulate complex likelihood landscapes in high energy physics and cosmology. This approach offers significant computational efficiency and improved resolution of confidence regions, especially for analyses with intricate correlations or degeneracies. The framework was successfully applied to study flavour anomalies in B meson decays and is adaptable to other phenomenological systems. Furthermore, SHAP values were employed to ensure the model's predictions are transparent and physically interpretable. AI
IMPACT This research demonstrates how ML can accelerate and improve the interpretability of complex scientific simulations, potentially speeding up discovery in fields like particle physics and cosmology.
RANK_REASON Academic paper detailing a new ML methodology for physics simulations. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
