Application of Machine Learning Methods for Detecting Atypical Structures in Astronomical Maps

TL;DR

This paper applies machine learning techniques to identify and analyze unusual structures in astronomical maps, specifically using Planck mission data to detect anomalies in cosmic microwave background radiation.

Contribution

It introduces a machine learning approach for detecting atypical structures in astronomical maps and provides a new algorithm for anomaly detection in cosmic microwave background data.

Findings

Detected multiple anomalous structures in Planck maps

Compiled a map of positions of atypical objects

Compared anomalies with known astrophysical phenomena

Abstract

The paper explores the use of various machine learning methods to search for heterogeneous or atypical structures on astronomical maps. The study was conducted on the maps of the cosmic microwave background radiation from the Planck mission obtained at various frequencies. The algorithm used found a number of atypical anomalous structures in the actual maps of the Planck mission. This paper details the machine learning model used and the algorithm for detecting anomalous structures. A map of the position of such objects has been compiled. The results were compared with known astrophysical processes or objects. Future research involves expanding the dataset and applying various algorithms to improve the detection and classification of outliers.