# Classification of Galaxy Cluster Membership with Machine Learning

**Authors:** Daniel Farid, Han Aung, Daisuke Nagai, Arya Farahi, Eduardo Rozo  
**arXiv:** [2205.01700](https://arxiv.org/abs/2205.01700)  
**Date:** 2022-05-03  
**Summary page:** https://www.summarizepaper.com/en/arxiv-id/2205.01700v1/

## Key points

- Authors present a classification algorithm using Random Forest machine learning model
- Algorithm differentiates galaxies into three populations: orbiting, infalling, and background
- Phase space information used as input for training and testing
- Algorithm accurately recovers distribution of orbiting and infalling galaxies with less than 1% error
- Probabilistic predictions used in presence of interlopers in projected phase space
- Machine learning-based galaxy classification method measures dynamics of cluster galaxies at percent level accuracy
- Potential applications discussed include improving cluster cosmology and understanding galaxy quenching
- Accurate classification helps gain insights into formation, evolution, and processes involved in galaxy quenching within clusters

## Summary

In their paper titled "Classification of Galaxy Cluster Membership with Machine Learning," authors Daniel Farid, Han Aung, Daisuke Nagai, Arya Farahi, and Eduardo Rozo present a classification algorithm that utilizes the Random Forest machine learning model to differentiate galaxies into three populations: orbiting, infalling, and background (interloper). The algorithm takes phase space information as input and is trained and tested using galaxies from the UniverseMachine mock catalog based on Multi-Dark Planck 2 N-body simulations. The authors demonstrate that their algorithm can accurately recover the distribution of orbiting and infalling galaxies in both 3D and 2D position and velocity distributions with less than 1% error. This is achieved by using probabilistic predictions in the presence of interlopers in the projected phase space. By employing this machine learning-based galaxy classification method, they are able to measure the dynamics of cluster galaxies at a percent level accuracy. The potential applications of this technique are discussed in the paper, particularly its ability to improve cluster cosmology and galaxy quenching. With more accurate classification of galaxy cluster membership, researchers can gain insights into the formation and evolution of clusters as well as understand the processes involved in galaxy quenching. Overall, this study presents a novel approach for classifying galaxies within clusters using machine learning algorithms. The results demonstrate its effectiveness in accurately identifying different populations of galaxies based on phase space information. This advancement has significant implications for our understanding of galaxy clusters and their dynamics; it can be used to improve our knowledge about cluster cosmology as well as provide further insight into how galaxies become quenched over time.