Identifying Exoplanets with Deep Learning III: Automated Triage and Vetting of TESS Candidates

Abstract

NASA's Transiting Exoplanet Survey Satellite (TESS) presents us with an unprecedented volume of space-based photometric observations that must be analyzed in an efficient and unbiased manner. With at least $\sim 1,000,000$ new light curves generated every month from full frame images alone, automated planet candidate identification has become an attractive alternative to human vetting. Here we present a deep learning model capable of performing triage and vetting on TESS candidates. Our model is modified from an existing neural network designed to automatically classify Kepler candidates, and is the first neural network to be trained and tested on real TESS data. In triage mode, our model can distinguish transit-like signals (planet candidates and eclipsing binaries) from stellar variability and instrumental noise with an average precision (the weighted mean of precisions over all classification thresholds) of 97.0% and an accuracy of 97.4%. In vetting mode, the model is trained to identify only planet candidates with the help of newly added scientific domain knowledge, and achieves an average precision of 69.3% and an accuracy of 97.8%. We apply our model on new data from Sector 6, and present 288 new signals that received the highest scores in triage and vetting and were also identified as planet candidates by human vetters. We also provide a homogeneously classified set of TESS candidates suitable for future training.

arXiv comment15 pages, 6 figures, 2 tables, accepted for publication in AJ

References in corpus (36)

• The K2 Mission: Characterization and Early results
• The WASP Project and the SuperWASP Cameras
• Kepler Mission Design, Realized Photometric Performance, and Early Science
• A box-fitting algorithm in the search for periodic transits
• The CoRoT satellite in flight : description and performance
• A Technique for Extracting Highly Precise Photometry for the Two-Wheeled Kepler Mission
• Wide-field millimagnitude photometry with HAT: A Tool for Extra-solar Planet Detection
• The Transiting Exoplanet Survey Satellite: Simulations of planet detections and astrophysical false positives
• TrES-1: The Transiting Planet of a Bright K0V Star
• A disintegrating minor planet transiting a white dwarf
• Identifying Exoplanets with Deep Learning: A Five Planet Resonant Chain around Kepler-80 and an Eighth Planet around Kepler-90
• Planet Hunters X. KIC 8462852 - Where's the Flux?
• Planetary Candidates Observed by Kepler. VII. The First Fully Uniform Catalog Based on The Entire 48 Month Dataset (Q1-Q17 DR24)
• Planetary Candidates from the First Year of the K2 Mission
• 197 Candidates and 104 Validated Planets in K2's First Five Fields
• TESS Discovery of a Transiting Super-Earth in the $\pi$ Mensae System
• 275 Candidates and 149 Validated Planets Orbiting Bright Stars in K2 Campaigns 0-10
• TESS Discovery of an ultra-short-period planet around the nearby M dwarf LHS 3844
• Searching for Exoplanets Using Artificial Intelligence
• Automatic Classification of Kepler Planetary Transit Candidates
• TESS full orbital phase curve of the WASP-18b system
• The EBLM Project IV. Spectroscopic orbits of over 100 eclipsing M dwarfs masquerading as transiting hot-Jupiters
• The KELT Follow-Up Network and Transit False Positive Catalog: Pre-vetted False Positives for TESS
• Scientific Domain Knowledge Improves Exoplanet Transit Classification with Deep Learning
• 60 Validated Planets from K2 Campaigns 5-8
• 44 Validated Planets from K2 Campaign 10
• Identifying Exoplanets with Deep Learning II: Two New Super-Earths Uncovered by a Neural Network in K2 Data
• Machine-learning Approaches to Exoplanet Transit Detection and Candidate Validation in Wide-field Ground-based Surveys
• A Machine Learning Technique to Identify Transit Shaped Signals
• An Eccentric Massive Jupiter Orbiting a Sub-Giant on a 9.5 Day Period Discovered in the Transiting Exoplanet Survey Satellite Full Frame Images
• Shallow Transits - Deep Learning I: Feasibility Study of Deep Learning to Detect Periodic Transits of Exoplanets
• Transit Shapes and Self Organising Maps as a Tool for Ranking Planetary Candidates: Application to Kepler and K2
• Identifying False Alarms in the Kepler Planet Candidate Catalog
• Planetary Candidates from K2 Campaign 16
• SIDRA: a blind algorithm for signal detection in photometric surveys
• A TESS Dress Rehearsal: Planetary Candidates and Variables from K2 Campaign 17

Cited by in corpus (26)

• The TESS Objects of Interest Catalog from the TESS Prime Mission
• ExoMiner: A Highly Accurate and Explainable Deep Learning Classifier that Validates 301 New Exoplanets
• The TESS Faint Star Search: 1,617 TOIs from the TESS Primary Mission
• Exoplanet Validation with Machine Learning: 50 new validated Kepler planets
• Warm Jupiters in TESS Full-Frame Images: A Catalog and Observed Eccentricity Distribution for Year 1
• The TESS Grand Unified Hot Jupiter Survey. II. Twenty New Giant Planets
• Classifying Exoplanet Candidates with Convolutional Neural Networks: Application to the Next Generation Transit Survey
• A Habitable-Zone Earth-Sized Planet Rescued from False Positive Status
• Identifying Exoplanets with Deep Learning. V. Improved Light Curve Classification for TESS Full Frame Image Observations
• Automated identification of transiting exoplanet candidates in NASA Transiting Exoplanets Survey Satellite (TESS) data with machine learning methods
• Identify Light-Curve Signals with Deep Learning Based Object Detection Algorithm. I. Transit Detection
• Nigraha: Machine-learning based pipeline to identify and evaluate planet candidates from TESS
• Identifying Light-curve Signals with a Deep Learning Based Object Detection Algorithm. II. A General Light Curve Classification Framework
• The TESS Triple-9 Catalog II: a new set of 999 uniformly-vetted exoplanet candidates
• Distinguishing a planetary transit from false positives: a Transformer-based classification for planetary transit signals
• DIAmante TESS AutoRegressive Planet Search (DTARPS): I. Analysis of 0.9 Million Light Curves
• Identifying Exoplanets with Deep Learning. IV. Removing Stellar Activity Signals from Radial Velocity Measurements Using Neural Networks
• The GPU Phase Folding and Deep Learning Method for Detecting Exoplanet Transits
• Identifying Potential Exomoon Signals with Convolutional Neural Networks
• Searching for Possible Exoplanet Transits from BRITE Data through a Machine Learning Technique
• ExoplANNET: A deep learning algorithm to detect and identify planetary signals in radial velocity data
• DIAmante TESS AutoRegressive Planet Search (DTARPS): II. Hundreds of New TESS Candidate Exoplanets
• Exoplanet Transit Candidate Identification in TESS Full-Frame Images via a Transformer-Based Algorithm
• DART-Vetter: A Deep LeARning Tool for automatic triage of exoplanet candidates
• Shallow Transits -- Deep Learning II: Identify Individual Exoplanetary Transits in Red Noise using Deep Learning
• Exploring Late Stellar Evolution in the Era of Large Surveys: Machine Learning Prospects for Hot Subdwarfs and White Dwarfs