search

hagabbar / matching_cnn_paper

Matching matched filtering with deep learning in gravitational-wave astronomy
2 stars 1 forks source link

Final plots should have better SNR resolution #28

Open chrismessenger opened 6 years ago

chrismessenger commented 6 years ago

When we converge on the network and datasets to use we should make nice smooth curves.

I propose that we make new datasets at SNR spacing = 1 and then fit smooth polynomials through them.

We do this at the very end when we are happy with the results.