A logistic regression model to identify possible pulsar candidates using PyTorch The model uses data from HTRU dataset with possible pulsar candidates to train our model on.
The implementation is based on pytorch library and functions as an initial experience of working on astronomical data using pytorch.
Talking about pulsars, these are ultra desne neutron stars which are rotating and giving out streams of radiation from its poles. We can only see them when the radiation is facing towards earth. Due to its rotation, we can only observe them for a fixed amount of time in one go, giving it a periodic observation or as we say, a Pulsating Waveform. Evidences show, detection of pulsar are especially diffcult, due to the fact that Neutron stars are not very luminous or big in size, thus other general methods don't work that efficiently for pulsars.