sakshambassi/deepstarclassification

Classification of variable stars using Convolutional Neural Networks

DeepStarClassification

Classification of variable stars using similar approach to Mahabal et al. The paper referred here is Deep-Learnt Classification of Light Curves( https://arxiv.org/pdf/1709.06257.pdf )

Procedure

This is a Research project that I am currently working on with Dr. Kaushal Sharma at Inter-University Centre for Astronomy and Astrophysics. We tried to synthesize images of star classes based on the difference in magnitude of brightness with respect to time period of it. These images are trained on Convolutional Neural Networks and binary classification of 2 classes is implemented.

Future Work

We have planned to implement Multi-class classification and use more parameters in CRTS data to improve our classification.

Acknowledgments

• Referred to Keras-team's code on Mnist_cnn.py (https://github.com/keras-team/keras/blob/master/examples/mnist_cnn.py)