Hunting Exoplanets Using Transit Method With Convolution Neural Networks


Finding planets which pose as potentially habitable earth-like planets outside our solar system are called as exoplanets. In 1995, the first such exoplanet found orbiting around a star in 4-day orbits. Further in 2009, NASA launched a spacecraft called Kepler to look for exoplanets wherein Kepler looked for planets in a wide range of sizes and orbits. Kepler observes the exoplanets using the “Transit Method”, that is when a planet moves in front of a star, there is a dip in the flux intensity (brightness) of the star. For this research project study, we used a binary Deep Learning model using TensorFlow after analyzing the wrangled data of the flux intensity variation. Thus, yielding an accurate model that outputs a binary value of 0 (or) 1 indicating True Positive [i.e. exoplanet detected] (or) True Negative [i.e. Not an exoplanet] which we found redeems the conclusive result of the system.



Software And Hardware

• Hardware: Processor: i3 ,i5 RAM: 4GB Hard disk: 16 GB • Software: operating System : Windws2000/XP/7/8/10 Anaconda,jupyter,spyder,flask Frontend :-python Backend:- MYSQL