Suppression of Continuum Background with Neural Networks for Belle II

Sumitted to PubDB: 2023-12-21

Category: Bachelor Thesis, Visibility: Public

Tags: -

Authors	Allen Caldwell, Hans-Günther Moser, Bela Urbschat
Date	Jan. 1, 2023
Belle II Number	BELLE2-UTHESIS-2023-002
Abstract	At Belle II, a so called B factory, pairs of B mesons are produced in large numbers to study their subsequent decays. B physics offers many opportunities for precision tests of the standard model, where processes involving loop diagrams are especially interesting as they may be influenced by new particles entering the loops. A group of decays of interest here are the B → Kπ decays, where the tree-level amplitudes are suppressed, thus making them sensitive to loop contributions. Measurements of branching ratios and CP asymmetries of those decays are expected to satisfy certain relations predicted by the standard model, which may however be violated if new physics is involved. This allows for so called null tests of the standard model. Some of the necessary measurements are however very difficult as the decays are rare and backgrounds are high. In this situation thus the best possible background suppression is desirable. Motivated by this, here a novel approach for qq background suppression using low level variables and deep neural networks will be explored for the B0 → K0 π 0 decay.
Conference	Munich