Access Type

Open Access Dissertation

Date of Award

January 2022

Degree Type


Degree Name



Physics and Astronomy

First Advisor

Cinabro David Anthony


Belle is a particle physics experiment based at the KEK laboratory in Tsukuba Japanwhich ran from 1999 to 2010 and collected 1ab−1 of data. The Belle experiment is focused on studying the properties of particles called B mesons which are produced by accelerating and colliding electron and positron beams. These B mesons show the biggest differences between the properties of matter and anti-matter of any known particles. One of the main goals of the Belle experiments is to understand the differences between matter and anti-matter, specifically violations of charge-parity symmetry (CP violation) and how anti-matter vanished and we come to live in a matter dominated universe. In this dissertation, I explore the charmless B decay B± → Ks0π±π0 with the Belle full Monti-Carlo (full MC) simulation and Belle data corresponding to 571f b−1 of luminosity and measure the decay’s branching fraction(BF). Charmless transitions can proceed by a b → u transition via a tree level diagram or b → s or d transition via the so-called penguin diagram. Both decay types are highly suppressed compared to the b → c transition and we expect a small branching fraction, smaller than 10−5. Penguin processes are important in B-meson charmless decays, thus it is possible to have contributions of unknown particles in the loop process leading to CP-violation contribution from Beyond Standard Model. Charmless B decays are sensitive to the angle γ (aka ϕ3) in the unitary triangle and open a new window to analyze the CP violation process.

I use the innovative Belle II software (basf2) for the initial reconstruction process. Thechallenge in observing the B± → Ks0π±π0 decay is to suppress backgrounds from continuum events, which do not contain b quarks, and background from other B meson decays. I use direct selections such as beam-energy constrained mass (Mbc), energy difference (∆E), probability of the vertex fit result (ChiProb), and others, on several variables to remove background from B± → Ks0π±π0 events. I find such an approach is insufficient. I found it was necessary to use a multi-variate analysis (MVA) machine learning/artificial intelligence technique called a boosted decision tree (BDT) to reduce the backgrounds to the level to allow me to clearly observe the decay and measure the BF. I compare my results with unpublished results from the BaBar experiment. Additionally I use the Dalitz plot (DP) technique to study the intermediate resonance contributions in this decay. I use the Laura++ software to generate and fit toy Monte Carlo(toy MC), full Monte Carlo simulated data, and, based on the techniques developed on these simulations, the experimental data to study the resonance sub-structure of this decay.