Poster abstract details

We make use of machine learning (ML) algorithms to try to correctly predict the masses of galaxy groups and clusters of galaxies from the Sloan Digital Sky Survey (SDSS). We use synthetic catalogues of galaxies that aim to represent the clustering and other properties of galaxies in the real Universe. We use Random Forest, XGBoost, and Neural Network to predict the masses of these galaxy groups, and compare our results with the more traditional methods, such as halo abundance matching (HAM) and dynamical masses. We find that ML algorithms tend to predict the group mass to within 10% of the true group masses. However, HAM esti- mates tend to be slightly more robust in determining high masses. We find that the luminosity of galaxies and (g − r) color are the most important features used when determining group and cluster masses.