Leveraging Reinforcement Learning, Genetic Algorithms and Transformers for background determination in particle physics

2025-09-18

AI Summary

In particle physics, identifying dominant backgrounds in beauty-hadron decays is challenging due to overwhelming combinatorial backgrounds with similar final states; current approaches rely heavily on expert knowledge and are computationally constrained, limiting systematic coverage of critical backgrounds. This work proposes a novel framework integrating reinforcement learning (RL), genetic algorithms (GA), and Transformer architectures: it is the first to jointly deploy RL and GA for decay-path search under sparse-reward conditions; introduces a Transformer to capture long-range dependencies in decay sequences; and enables automatic, scalable identification of key backgrounds across large decay spaces. Experiments demonstrate a substantial reduction in reliance on human priors, with significant improvements in both background classification accuracy and coverage. The framework exhibits strong generalizability to other high-energy physics measurement tasks.

Technology Category

• Machine Learning: Reinforcement Learning
• Search and Optimization: Learning to Search
• Reasoning under Uncertainty: Sequential Decision Making

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Experimental studies of beauty hadron decays face significant challenges due to a wide range of backgrounds arising from the numerous possible decay channels with similar final states. For a particular signal decay, the process for ascertaining the most relevant background processes necessitates a detailed analysis of final state particles, potential misidentifications, and kinematic overlaps, which, due to computational limitations, is restricted to the simulation of only the most relevant backgrounds. Moreover, this process typically relies on the physicist's intuition and expertise, as no systematic method exists. This paper has two primary goals. First, from a particle physics perspective, we present a novel approach that utilises Reinforcement Learning (RL) to overcome the aforementioned challenges by systematically determining the critical backgrounds affecting beauty hadron decay measurements. While beauty hadron physics serves as the case study in this work, the proposed strategy is broadly adaptable to other types of particle physics measurements. Second, from a Machine Learning perspective, we introduce a novel algorithm which exploits the synergy between RL and Genetic Algorithms (GAs) for environments with highly sparse rewards and a large trajectory space. This strategy leverages GAs to efficiently explore the trajectory space and identify successful trajectories, which are used to guide the RL agent's training. Our method also incorporates a transformer architecture for the RL agent to handle token sequences representing decays.

Problem

• Systematically determining critical backgrounds in beauty hadron decay measurements
• Overcoming sparse rewards and large trajectory spaces in reinforcement learning
• Handling token sequences representing particle decays using transformer architecture

Innovation

• Reinforcement Learning for background determination
• Genetic Algorithms explore sparse reward trajectories
• Transformer architecture processes decay token sequences