# Clustering of Hotspots in the Cosmic Microwave Background

## TL;DR

This paper applies clustering algorithms to analyze the distribution of hotspots in the CMB, revealing non-Poissonian patterns that suggest early universe structure formation, offering new insights into cosmology.

## Contribution

It introduces clustering-based statistical analysis of CMB hotspots, demonstrating non-Poissonian distributions and implications for early universe structure.

## Key findings

- Hotspot counts do not follow a Poisson distribution.
- Distribution fits better to Negative Binomial and GQED models.
- Evidence suggests early universe structure existed during recombination.

## Abstract

The physics behind the origin and composition of the Cosmic Microwave Background (CMB) is a well-established topic in the field of Cosmology. Literature on CMB anisotropies reveal consistency with Gaussianity, but these were conducted on full multi-frequency temperature maps. In this thesis, we utilise clustering algorithms to specifically conduct statistical analyses on the distribution of hotspots in the CMB. We describe a series of data processing and clustering methodologies conducted, with results that conclusively show that the counts-in-cells distribution of hotspots in the CMB does not follow a Poisson distribution. Rather, the distribution exhibits a much closer fit to both the Negative Binomial Distribution (NBD) and the Gravitational Quasi-Equilibrium Distribution (GQED). From this result, we conclude that structure likely existed in the early universe, from the period of the recombination Epoch, possibly opening new insights in the field of galaxy formation.