Ramon Winterhalder (University of Milano Statale)

Machine Learning for HEP - Just a trend or a paradigm shift?

Machine learning (ML) has rapidly evolved from a promising idea to a powerful tool in high-energy physics (HEP). In experimental analyses, ML-based methods have already demonstrated significant improvements in sensitivity, effectively reducing the amount of data required to achieve discovery-level results. However, the role of ML in HEP goes far beyond classification tasks. ML is also transforming theoretical predictions and event generation. With examples from the MadNIS framework for neural importance sampling and fast surrogate models for expensive color amplitudes, I will highlight how ML can accelerate Monte Carlo simulations and enable precision calculations that would otherwise be computationally prohibitive.