Abstract Details
| Name: Vaibhav Dixit Affiliation: Physical Research Laboratory, Ahmedabad Conference ID: ASI2018_1595 Title : Deep learning based models for the classification of exoplanetary emission spectra Authors and Co-Authors : Vaibhav Dixit Physical Research Laboratory Abstract Type : Poster Abstract Category : Stars,ISM and the Galaxy Abstract : Spectral retrieval of exoplanetary atmospheres often demands preselection of user-defined molecular/atomic opacities. This manual intervention introduces biases in the retrieval process. Moreover high dimensionality of retrieval parameters enhances the computational complexity. Models based on deep learning can overcome these challenges. Deep learning is a subset of Artificial Intelligence which takes inspiration from model of human brain. Deep learning encourages computational models that are built on multiple layers to learn representations of data with multiple levels of abstraction. In recent years Generative Adversarial Network(GAN) has emerged as powerful tool in deep learning. This work is aimed at introducing a GAN based model to accurately identify the molecular signatures for a wide variety of planets, compositions and atmospheric thermal profiles. |