Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Bayesian Active Search on Parameter Space: a 95 GeV Spin-0 Resonance in the ($B-L$)SSM

Mauricio A. Diaz, Giorgio Cerro, Srinandan Dasmahapatra, Stefano Moretti

Abstract

In the attempt to explain possible data anomalies from collider experiments in terms of New Physics (NP) models, computationally expensive scans over their parameter spaces are typically required in order to match theoretical predictions to experimental observations. Under the assumption that anomalies seen at a mass of about 95 GeV by the Large Electron-Positron (LEP) and Large Hadron Collider (LHC) experiments correspond to a NP signal, which we attempt to interpret as a spin-0 resonance in the $(B-L)$ Supersymmetric Standard Model ($(B-L)$SSM), chosen as an illustrative example, we introduce a novel Machine Learning (ML) approach based on a multi-objective active search method, called b-CASTOR, able to achieve high sample efficiency and diversity, due to the use of probabilistic surrogate models and a volume based search policy, outperforming competing algorithms, such as those based on Markov-Chain Monte Carlo (MCMC) methods.

Bayesian Active Search on Parameter Space: a 95 GeV Spin-0 Resonance in the ($B-L$)SSM

Abstract

In the attempt to explain possible data anomalies from collider experiments in terms of New Physics (NP) models, computationally expensive scans over their parameter spaces are typically required in order to match theoretical predictions to experimental observations. Under the assumption that anomalies seen at a mass of about 95 GeV by the Large Electron-Positron (LEP) and Large Hadron Collider (LHC) experiments correspond to a NP signal, which we attempt to interpret as a spin-0 resonance in the Supersymmetric Standard Model (SSM), chosen as an illustrative example, we introduce a novel Machine Learning (ML) approach based on a multi-objective active search method, called b-CASTOR, able to achieve high sample efficiency and diversity, due to the use of probabilistic surrogate models and a volume based search policy, outperforming competing algorithms, such as those based on Markov-Chain Monte Carlo (MCMC) methods.
Paper Structure (19 sections, 42 equations, 9 figures, 3 tables, 1 algorithm)

This paper contains 19 sections, 42 equations, 9 figures, 3 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Higgs Bosons in the $(B-L)$SSM
  3. Experimental Searches
  4. The (B-L)SSM
  5. ($B-L$)SSM Predictions and HEP Software
  6. Active Search Formulation
  7. Search Policy
  8. Constraint Active Search
  9. Batch Evaluation
  10. Surrogate Models
  11. b-CASTOR
  12. Benchmark Algorithm
  13. Technical Implementation
  14. Results
  15. Double-objective 2D Test Function
  16. ...and 4 more sections

Figures (9)

  • Figure 1: Ground truth for the 2D double objective test function $\mathbf{f}_{BH}(\theta_1, \theta_2)$ as per eq. \ref{['eq:testf']}, featuring contour levels to demonstrate the constraints on the objectives. On the left, the $\mathcal{S}$ region within the search space is depicted.
  • Figure 2: Performance metrics across ten independent runs for b-CASTOR (blue) and MCMC-MH (green), with mean values depicted in darker shades.
  • Figure 3: b-CASTOR results for different independent runs.
  • Figure 4: MCMC-MH results for different independent runs.
  • Figure 5: Performance metrics for b-CASTOR (blue) and MCMC-MH (orange), for the search in $\mathcal{H}_{{(B-L){\mathrm{SSM}}}}$ fitting $\mu_{\gamma\gamma}^\mathrm{exp}$.
  • ...and 4 more figures