The Four Cosmic Tidal Web Elements from the β-skeleton

Abstract Precise cosmic web classification of observed galaxies in massive spectroscopic surveys can be either highly uncertain or computationally expensive. As an alternative, we explore a fast Machine Learning-based approach to infer the underlying dark matter tidal environment galaxy distribution from its β -skeleton graph. We develop and test our methodology using cosmological magnetohydrodynamic simulation Illustris-TNG at z = 0. three different tree-based machine-learning algorithms find that random forest classifier best use graph-based features classify as belonging peak, filament, sheet defined by T-Web algorithm. The match between corresponds density field smoothed over scales 2 Mpc, threshold eigenvalues dimensionless tensor λ th 0.0, number densities around 8 × 10 −3 Mpc . This results on weighted F1 score 0.728 global accuracy 74%. More extensive tests take into account light-cone effects redshift space distortions are left for future work. make one highest ranking models available public repository reference reuse.