Powered by Indico v3.0.3
Abstract:
EXO-200 was an experiment searching the neutrinoless double beta decay using a liquid-phase time projection chamber filled with Xe-136. It accumulated a large low-background dataset of physics data, as well as unique calibration data that remain to be a perfect testbed for novel analysis algorithms, such as deep neural networks (DNNs). After introducing the EXO-200 experiment, this talk describes successful applications of DNNs to simulate, reconstruct, and analyze the EXO-200 data. We demonstrate that DNNs are able to extract necessary high-level features directly from raw waveforms, with minimal pre- processing. The developed DNNs and conventional analyses approaches are compared back-to-back. Importantly, the DNNs described in this talk are validated with real detector data, and some are even trained on the data. This shows the potential of DNNs in reducing or even eliminating the reliance on the conventional Monte Carlo and its imperfections. The talk concludes by discussing advantages and challenges of DNNs for the upcoming next-generation low-background experiments.
About the speaker:
Dr. Igor Ostrovskiy worked at JINR and INFN before receiving his Ph.D. at the University of Alabama, where he performed the first analysis of the Double Chooz far detector data that provided a hint of non-zero value of the neutrino mixing angle θ13. He then did a postdoc at Stanford before returning to the University of Alabama where he currently holds an associate professor position. His group's two primary research foci are searches for magnetic monopoles at the LHC and R&D for the next generation of liquid-xenon detectors that will search for dark mater and neutrinoless double beta decay. He is a member of the MoEDAL, DARWIN, EXO, and MONUMENT collaborations.
----------
会议号 Meeting ID: 9472 1492 781
会议链接 Meeting URL：https://zoom.us/j/94721492781?pwd=bVM0QWFnV3VTMnhqaytuRHE1cy9lUT09
入会密码Password：240314
主持密码Host key：068014