Planetary System Stability Data for Bayesian Neural Network Training

A dataset used for training and evaluating models in the paper "A Bayesian neural network predicts the dissolution of compact planetary systems." The data was created by Miles Cranmer of Princeton University and is associated with code on GitHub. The specific number of rows and features is not provided in the available metadata.

Use Cases

• Training Bayesian neural networks to predict planetary system stability based on orbital parameters.
• Benchmarking chaos prediction models for compact multi-planet systems.
• Studying the relationship between initial orbital configurations and system dissolution timescales.

Strengths

• Dataset is directly linked to a peer-reviewed arXiv paper (2101.04117), providing academic context.
• Associated code for working with the dataset is publicly available on GitHub, facilitating reproducibility.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment for large-scale modeling.
• Last update date is unknown; freshness unverified.

Provenance

Source

Miles Cranmer, Princeton University

Collection Method

Likely contains simulated or observational data on planetary system parameters for chaos modeling.