Identifying Exoplanets with Deep Learning: A Five-planet Resonant Chain around Kepler-80 and an Eighth Planet around Kepler-90
Problem
Weak Kepler signals at low signal-to-noise are hard to distinguish from false positives, blocking accurate occurrence rates of Earth-sized planets.
Why it matters
Determining how common Earth-sized planets are requires reliably vetting faint candidate signals.
What the AI did
A deep convolutional neural network was trained to classify signals as transiting exoplanets or false positives and rank candidates by planet likelihood.
Results
Ranked plausible planet signals above false positives 98.8% of the time on test set; per title, identified a five-planet resonant chain around Kepler-80 and an eighth planet around Kepler-90.
Impact
Enabled automated vetting of weak candidates and discovery of new multi-planet systems around Kepler-80 and Kepler-90.
Record
| Venue | The Astronomical Journal |
|---|---|
| Published | 2018 |
| Work type | article |
| Authors | Christopher J. Shallue, Andrew M. Vanderburg |
| Citations | 368 (OpenAlex, at import time) |
| References | 96 |
| Model or method | deep convolutional neural network |
| Organization | not stated in source |
| Primary source | https://doi.org/10.3847/1538-3881/aa9e09 |
| Open access | hybrid |
| Keywords | deep learning, exoplanets, Kepler, false positives, convolutional neural network |
Importance rating
real but narrower advance
Abstract states determining occurrence rates of Earth-sized planets requires reliable vetting at low signal-to-noise; significant for exoplanet science but a narrow research domain.
More in Physics & Astronomy
| # | Paper | Year | Citations | Importance |
|---|---|---|---|---|
| 1 | Magnetic control of tokamak plasmas through deep reinforcement learning | 2022 | 739 |
7
|
| 2 | Solving the quantum many-body problem with artificial neural networks | 2017 | 2,205 |
7
|
| 3 | Learning high-accuracy error decoding for quantum processors | 2024 | 90 |
7
|
| 4 | Optimizing Quantum Error Correction Codes with Reinforcement Learning | 2019 | 158 |
7
|
| 5 | Machine-learning-assisted correction of correlated qubit errors in a topological code | 2018 | 141 |
7
|