Artificial Intelligence and Machine Learning for Simulation-Based Inference in $B^0 \to K^{*0} \ell^+ \ell^-$

Sumitted to PubDB: 2025-12-11

Category: Master Thesis

Tags:

Principal Authors	Ethan Lee
Date	2025-12-11
Belle II Number	BELLE2-MTHESIS-2025-032
Abstract	We investigate neural simulation-based inference approaches to fitting the deviation of Wilson Coefficient 9 ($C_9$) from its Standard Model value ($C_9^{SM}$) given ${B^0 \to K^{*0} \mu^+ \mu^-}$ events simulated in the context of the Belle II experiment. We denote this deviation as $\delta C_9$. We compare three neural network-based approaches to this multi-dimensional fitting problem. The first approach converts the dataset into a three-dimensional grid and fits for $\delta C_9$ using computer vision techniques. The second approach uses the deep sets architecture to predict $\delta C_9$ from a dataset while enforcing the permutation invariance of events. The third approach trains a classification model to predict a binned probability distribution over $\delta C_9$ given a single event. Predictions are then aggregated using the independence of events to obtain the binned $\delta C_9$ probability distribution given the entire dataset. We train and evaluate models on simulated datasets with and without detector effects. We also train and evaluate models on a dataset that includes simulated background events from the $M_{bc}$ sideband.
Institute	Hawaii