Image Details

Choose export citation format:

Machine Learning for Radial Velocity Analysis. I. Vision Transformers as a Robust Alternative for Detecting Planetary Candidates

  • Authors: Anoop Gavankar, Tanish Mittal, Joe P. Ninan, Shravan Hanasoge

Anoop Gavankar et al 2026 The Astronomical Journal 171 .

  • Provider: AAS Journals

Caption: Figure 31.

Distribution of predicted semi-amplitude classes for systems without injected planetary signal (the “no planet” scenario). Amplitudes are expressed in m s−1. The predictions are heavily skewed toward the lowest amplitude bin, with over 80% of samples assigned to this class. While the remaining predictions are primarily assigned to the correct no planet class, the overall classification accuracy remains limited to approximately 20%. This reflects the model’s tendency to infer low-amplitude planetary signals even when none are present. This highlights the model’s tendency to predict low amplitudes in the absence of a true planetary signal.

Other Images in This Article

Show More

Copyright and Terms & Conditions

Additional terms of reuse