Skip to main content
All Reviews
Astronomy
intermediate

Artificial neural network classification of asteroids in the M1:2 mean-motion resonance with Mars

V Carruba et al. (2021)

Published
Apr 7, 2021
Journal
Monthly Notices of the Royal Astronomical Society · Vol. 504 · No. 1
DOI
10.1093/mnras/stab914

At a GlanceAI

First ANN-based classifier to automatically identify asteroid orbital behavior in Mars’ M1:2 mean-motion resonance.

SummaryAI

The study introduces (for the first time) artificial neural networks to automatically classify asteroid orbital behavior associated with the Mars M1:2 mean-motion resonance. It reports >85% performance on classifying images of resonant-argument behavior, enabling classification of all numbered asteroids in the resonance region and supervised predictions for multi-opposition objects. The results support that the M1:2 resonance primarily influences members of the Massalia, Nysa, and Vesta families, providing a scalable way to map resonance-driven dynamics across large asteroid samples.

Method SnapshotAI

Supervised neural-network image classification of resonant-argument behavior, with model optimization using genetic algorithms.

BackgroundAI

Mean-motion resonance dynamics in the asteroid belt and basic supervised machine learning (neural networks and evaluation metrics).

Expert Review: Artificial neural network classification of asteroids in the M1:2 mean-motion resonance with Mars | Marginalia