| 研究実績の概要 |
This research aimed to predict the inflated radius of exoplanets using simulations. Initial results were disappointing, so we developed a neural network to identify patterns in the catalogue of exoplanet properties. Neural networks excel at finding trends in high-dimensional data that cannot be plotted on a graph.
While neural networks cannot reveal the association between properties, the trained network could estimate missing planetary properties, such as mass and radius. It is frequently impossible (even with further observations) to measure these properties, making these estimates valuable tools in developing planet formation theories.
This work also included a published outreach and education tool for exploring how the Earth’s environment is effected by small changes in properties.
|
