The Kepler and TESS missions have generated over 100,000 potential transit signals that must be processed in order to create a catalog of planet candidates. In recent years, there has been a growing interest in using machine learning to analyze this data in search of new exoplanets. However, existing machine learning works do not mimic how domain experts examine diagnostic tests to vet a transit signal. This is where ExoMiner comes into play - it is a highly accurate, explainable, and robust deep learning classifier that mimics the manual vetting process used by experts. ExoMiner allows for the validation of 301 new exoplanets from the MAST Kepler Archive and is general enough to be applied across missions such as the ongoing TESS mission. The modular design of ExoMiner favors its explainability. A simple explainability framework provides experts with feedback on why ExoMiner classifies a transit signal into a specific class label (e.g., planet candidate or not planet candidate). Combined with flux data in multiple forms, these diagnostic tests are the main components of the 1-page DV summary and the longer DV report used by experts to vet transit signals. The 1-page DV summary report includes stellar parameters, unfolded flux data, phase-folded full-orbit and transit view of the flux data, phase-folded transit view of the secondary eclipse, phase-folded whitened transit view, phase-folded odd and even transit view, difference image to check for false positives due to a background object being the source of the transit signature, and an analysis table that provides values related to the model fit and various DV diagnostic parameters. The analysis performed in this paper will allow for future directions for improving and building upon this work. First, changes will be made to the ExoMiner architecture to include new important features such as the number of transits and uncertainty about the depth of odd and even views to ensure that the model's use of these branches is more effective. Second, a full explainability study will be performed by utilizing more advanced techniques and designing a framework that can provide experts with insight about ExoMiner's labels and scores. Thirdly there will be more depth analysis of how label noise affects underrepresented scenarios. In conclusion, ExoMiner is a highly accurate and explainable deep learning classifier that mimics how domain experts examine diagnostic tests to vet a transit signal. It allows for validation of new exoplanets from Kepler Archive and can be applied across missions such as TESS. Future directions include changes in architecture to include new important features; performing a full explainability study utilizing more advanced techniques; analyzing label noise; exploring underrepresented scenarios in more depth; adapting ExoMiner for TESS data; providing feedback on why ExoMiner classifies signals into specific classes; incorporating stellar parameters; examining unfolded flux data; studying phase-folded full orbit views; assessing secondary eclipses; investigating whitened transits views; checking odd/even views for false positives due to background objects; analyzing model fits/DV diagnostics parameters etc..
- Kepler and TESS missions have generated over 100,000 potential transit signals that must be processed to create a catalog of planet candidates
- Machine learning is being used to analyze this data in search of new exoplanets
- Existing machine learning works do not mimic how domain experts examine diagnostic tests to vet a transit signal
- ExoMiner is a highly accurate, explainable, and robust deep learning classifier that mimics the manual vetting process used by experts
- ExoMiner allows for the validation of 301 new exoplanets from the MAST Kepler Archive and can be applied across missions such as TESS
- The modular design of ExoMiner favors its explainability with a simple framework providing feedback on why ExoMiner classifies a transit signal into a specific class label (e.g., planet candidate or not planet candidate)
- Diagnostic tests are the main components of the 1-page DV summary and longer DV report used by experts to vet transit signals.
- Future directions include changes in architecture to include new important features; performing a full explainability study utilizing more advanced techniques; analyzing label noise; exploring underrepresented scenarios in more depth; adapting ExoMiner for TESS data; providing feedback on why ExoMiner classifies signals into specific classes; incorporating stellar parameters; examining unfolded flux data; studying phase-folded full orbit views; assessing secondary eclipses; investigating whitened transits views; checking odd/even views for false positives due to background objects; analyzing model fits/DV diagnostics parameters etc.
Scientists are using machines to find new planets in space. They have a lot of data to look through, so they made a special machine called ExoMiner to help them. ExoMiner is very good at finding new planets and can even explain why it thinks something is a planet or not. Experts use special tests to make sure the signals from space are really planets, and ExoMiner works like those tests. In the future, scientists will make ExoMiner even better by adding more features and studying different views of the data.
Definitions
- Transit signals: A signal that occurs when a planet passes in front of its star.
- Machine learning: The process of teaching computers to learn and improve on their own without being explicitly programmed.
- Exoplanets: Planets that orbit stars outside our solar system.
- Deep learning classifier: A type of machine learning algorithm that can classify information based on patterns and features.
- Modular design: A design approach where a system is broken down into smaller parts (modules) that can be easily modified or replaced.
- Explainability: The ability for a machine learning model to provide explanations for its decisions or predictions.
- Diagnostic tests: Tests used by experts to verify if transit signals are actually caused by planets or something else.
- Label noise: Errors in labeling data that can affect the accuracy of machine learning models.
- Stellar parameters: Characteristics of stars such as size, temperature, and brightness.
- Flux data: Measurements of light
Exploring ExoMiner: A Highly Accurate and Explainable Deep Learning Classifier for Vetting Transit Signals
The Kepler and TESS missions have generated over 100,000 potential transit signals that must be processed in order to create a catalog of planet candidates. In recent years, there has been a growing interest in using machine learning to analyze this data in search of new exoplanets. However, existing machine learning works do not mimic how domain experts examine diagnostic tests to vet a transit signal. This is where ExoMiner comes into play - it is a highly accurate, explainable, and robust deep learning classifier that mimics the manual vetting process used by experts.
What Is ExoMiner?
ExoMiner is an advanced deep learning classifier designed to accurately identify exoplanet candidates from the MAST Kepler Archive and other mission data such as TESS. It uses flux data in multiple forms combined with diagnostic tests which are the main components of the 1-page DV summary report used by experts to vet transit signals. The 1-page DV summary includes stellar parameters, unfolded flux data, phase-folded full-orbit and transit view of the flux data, phase-folded transit view of the secondary eclipse, phase-folded whitened transit view, phase-folded odd and even transit view etc., along with an analysis table providing values related to model fit and various DV diagnostic parameters.
How Does ExoMiner Work?
ExoMiner's modular design favors its explainability; it provides feedback on why it classifies a given signal into specific classes (e.g., planet candidate or not planet candidate). It also allows for validation of 301 new exoplanets from MAST Kepler Archive while being general enough to be applied across missions such as TESS mission too. The accuracy achieved by ExoMiner surpasses all existing methods due its ability to mimic expert manual vetting processes closely while still being able to detect subtle differences between true planets versus false positives caused by background objects or other sources of noise present in light curves (a graph showing changes in brightness over time).
Future Directions For Improving And Building Upon This Work
ExoMiner can be further improved upon by making changes in architecture such as including more important features like number of transits/uncertainty about depth etc., performing full explainability study utilizing more advanced techniques & designing framework that can provide insight about labels & scores assigned by Exomineer; analyzing label noise; exploring underrepresented scenarios in more depth; adapting Exomineer for TESS data etc.. All these steps will help make sure that model’s use of branches is more effective & accurate when detecting true planets from false positives caused due background objects or other sources present in light curves
In conclusion, Exomineer is an extremely useful tool when it comes to identifying exoplanet candidates quickly & accurately while still maintaining high levels of explainability so domain experts can understand why certain decisions were made during classification process without having access complex algorithms behind them
