We introduce SuperNNova, an open source supernova photometric classification framework which leverages recent advances in deep neural networks. Our core algorithm is a recurrent neural network (RNN) that is trained to classify light-curves using photometric information only. Additional information such as host-galaxy redshift can be incorporated to improve performance. We evaluate our framework using realistic supernovae simulations that include survey detection. We show that our method, for the type Ia vs. non Ia supernovae classification problem, reaches accuracies greater than 96.92 ± 0.09 without any redshift information and up to 99.55 ± 0.06 when redshift, either photometric or spectroscopic, is available. Further, we show that our method attains unprecedented performance for classification of incomplete light-curves, reaching accuracies > 86.4 ± 0.1 (> 93.5 ± 0.8) without host-galaxy redshift (with redshift information) two days before maximum light. In contrast with previous methods, there is no need for time-consuming feature engineering and we show that our method scales to very large datasets with a modest computing budget. In addition, we investigate often neglected pitfalls of machine learning algorithms. We show that commonly used algorithms suffer from poor calibration and overconfidence on out-of-distribution samples when applied to supernovae data. We devise extensive tests to estimate the robustness of classifiers and cast the learning procedure under a Bayesian light, demonstrating a much better handling of uncertainties. We study the benefits of Bayesian RNNs for SN Ia cosmology. Our code is open-sourced and available in github ‡.
A : We present the results of measurements demonstrating the efficiency of the EDELWEISS-III array of cryogenic germanium detectors for direct dark matter searches. The experimental setup and the FID (Fully Inter-Digitized) detector array is described, as well as the efficiency of the double measurement of heat and ionization signals in background rejection. For the whole set of 24 FID detectors used for coincidence studies, the baseline resolutions for the fiducial ionization energy are mainly below 0.7 keV ee (FHWM) whereas the baseline resolutions for heat energies are mainly below 1.5 keV ee (FWHM). The response to nuclear recoils as well as the very good discrimination capability of the FID design has been measured with an AmBe source. The surface βand α-decay rejection power of R surf < 4 × 10 −5 per α at 90% C.L. has been determined with a 210 Pb source, the rejection of bulk γ-ray events has been demonstrated using γ-calibrations with 133 Ba sources leading to a value of R γ−mis−fid < 2.5 × 10 −6 at 90% C.L.. The current levels of natural radioactivity measured in the detector array are shown as the rate of single γ background. The fiducial volume fraction of the FID detectors has been measured to a weighted average value of (74.6 ± 0.4)% using the cosmogenic activation of the 65 Zn and 68,71 Ge isotopes. The stability and uniformity of the detector response is also discussed. The achieved resolutions, thresholds and background levels of the upgraded EDELWEISS-III detectors in their setup are thus well suited to the direct search of WIMP dark matter over a large mass range.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.