Can a machine learn the outcome of planetary collisions?

Abstract

Planetary-scale collisions are common during the last stages of formation of solid planets, including the Solar system terrestrial planets. The problem of growing planets has been divided into studying the gravitational interaction of embryos relevant in million year timescales and treated with N-body codes and the collision between objects with a timescale of hours to days and treated with smoothed-particle hydrodynamics. These are now being coupled with simple parameterized models. We set out to investigate if machine learning techniques can offer a better solution by predicting the outcome of collisions which can then be used in N-body simulations. We considered three different supervised machine learning approaches: gradient boosting regression trees, nested models, and gaussian processes. We found that the former produced the best results, and that it was slightly surpassed by ensembling different algorithms. With gaussian processes, we found the regions of parameter space that may yield the most information to machine learning algorithms. Thus, we suggest SPH calculations to focus first on mass ratios above 0.5.

arXiv comment13 pages, 11 figure, 2 tables. Submitted to ApJ

References in corpus (14)

• XGBoost: A Scalable Tree Boosting System
• Making other Earths: Dynamical Simulations of Terrestrial Planet Formation and Water Delivery
• Building the Terrestrial Planets: Constrained Accretion in the Inner Solar System
• Collisions Between Gravity-Dominated Bodies: 1. Outcome Regimes and Scaling Laws
• Collisions between Gravity-Dominated Bodies: 2. The Diversity of Impact Outcomes during the End Stage of Planet Formation
• Migration and the formation of systems of hot super-Earths and Neptunes
• The Compositional Diversity of Extrasolar Terrestrial Planets: I. In-Situ Simulations
• Terrestrial Planet Formation I. The Transition from Oligarchic Growth to Chaotic Growth
• Minimum Radii of Super-Earths: Constraints from Giant Impacts
• Formation of Terrestrial Planets from Protoplanets under a Realistic Accretion Condition
• Collisional Stripping and Disruption of Super-Earths
• Water/Icy Super-Earths: Giant Impacts and Maximum Water Content
• Realistic On-The-Fly Outcomes of Planetary Collisions: Machine Learning Applied to Simulations of Giant Impacts
• The dynamical evolution of transiting planetary systems including a realistic collision prescription

Cited by in corpus (6)

• Augmenting astrophysical scaling relations with machine learning: application to reducing the Sunyaev-Zeldovich flux-mass scatter
• The Role of Giant Impacts in Planet Formation
• Predicting the redshift of gamma-ray loud AGNs using supervised machine learning
• Machine learning applied to simulations of collisions between rotating, differentiated planets
• Residual Neural Networks for the Prediction of Planetary Collision Outcomes
• Random Forests applied to High Precision Photometry Analysis with Spitzer IRAC