BEGIN:VCALENDAR VERSION:2.0 X-WR-CALNAME:EventsCalendar PRODID:-//hacksw/handcal//NONSGML v1.0//EN CALSCALE:GREGORIAN BEGIN:VTIMEZONE TZID:America/New_York LAST-MODIFIED:20240422T053451Z TZURL:https://www.tzurl.org/zoneinfo-outlook/America/New_York X-LIC-LOCATION:America/New_York BEGIN:DAYLIGHT TZNAME:EDT TZOFFSETFROM:-0500 TZOFFSETTO:-0400 DTSTART:19700308T020000 RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU END:DAYLIGHT BEGIN:STANDARD TZNAME:EST TZOFFSETFROM:-0400 TZOFFSETTO:-0500 DTSTART:19701101T020000 RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU END:STANDARD END:VTIMEZONE BEGIN:VEVENT CATEGORIES:College of Engineering DESCRIPTION:Abstract:      Gravitational-wave detectors, such as the Las er Interferometer Gravitational-Wave Observatory (LIGO), are highly sensit ive instruments designed to detect weak astrophysical signals, requiring s ophisticated data analysis techniques to distinguish them from noise. This proposal aims to develop machine learning and data-driven methods to addr ess challenges in gravitational-wave astronomy, with applications to signa l detection, detector characterization, and low-latency analyses. We first address the detection of precessing binary black hole (BBH) systems. Beca use a fully precessing search is computationally prohibitive, current sear ches filter with aligned-spin template banks and are not optimally sensiti ve to precessing BBHs. We have performed aligned-spin searches using the S GNL pipeline on data containing simulated aligned-spin and precessing BBH signals, producing matched-filter triggers for both classes. We propose to train a feed-forward neural network on the recovered trigger parameters, such as the signal-to-noise ratio, consistency between the data and the be st-fitting template, and recovered masses and spins, to identify triggers likely to arise from precessing signals. This classifier will serve as a p reprocessing step to rapidly identify the small subset of triggers that re quire further follow-up. For this subset, we propose an optimization metho d that combines the signal-to-noise ratio with the autocorrelation-based l east-squares signal consistency test to improve the ranking of precessing candidates. We next address the characterization of transient bursts of de tector noise, or “glitches.” The LIGO detectors record data from more than 200,000 auxiliary channels that monitor the instrument and its enviro nment. We have developed a GPU-accelerated framework to transform time-ser ies data from these auxiliary channels into the fractal dimension, a compa ct representation of signal complexity. These representations will be used to train an unsupervised autoencoder to learn nominal detector behavior a nd flag anomalous activity. We will use the detected anomalies to identify auxiliary channels that witness distinct glitch populations, pointing to their environmental or instrumental origin. Finally, we address the need f or low-latency gravitational-wave alerts for multimessenger astronomy. Pub lic alerts are currently issued approximately 30 seconds after a candidate is identified using GWCelery, a Python framework built on the Celery dist ributed task queue, in which each stage records its result in the GraceDB event database before the next proceeds. As part of the LIGO Scientific Co llaboration, we are developing SGN-LLAI, a streaming infrastructure built on the Stream Graph Navigator (SGN) library, in which results pass directl y between stages via Kafka topics and are asynchronously recorded in Grace DB, removing database operations from the critical path. SGN-LLAI currentl y produces preliminary alerts in the standard public format, with latencie s as low as 5 seconds in initial tests. The proposed next phase will compl ete feature parity with GWCelery by adding source classification and prope rties, data quality information, and the full progression of public alerts from preliminary to initial and update notices, including human vetting a nd retraction. Alerts will then be disseminated through the GCN and SCiMMA networks. Advisor(s): Dr. Sarah Caudill, Department of Physics (scaudill@ umassd.edu)                                                      Committee members: Dr. Scott Field, Department of Mathe matics Dr. Vijay Varma, Department of Mathematics Dr. Derek Davis, Departm ent of Physics – University of Rhode Island Note: All EAS Students are encouraged to attend.\nEvent page: https://www.umassd.edu/events/cms/10-22 -26-eas-doctoral-proposal-defense-by-anushka-doke.php X-ALT-DESC;FMTTYPE=text/html:

Abstract:     

\n

Gravi tational-wave detectors\, such as the Laser Interferometer Gravitational-W ave Observatory (LIGO)\, are highly sensitive instruments designed to dete ct weak astrophysical signals\, requiring sophisticated data analysis tech niques to distinguish them from noise. This proposal aims to develop machi ne learning and data-driven methods to address challenges in gravitational -wave astronomy\, with applications to signal detection\, detector charact erization\, and low-latency analyses.

\n

We first address the detecti on of precessing binary black hole (BBH) systems. Because a fully precessi ng search is computationally prohibitive\, current searches filter with al igned-spin template banks and are not optimally sensitive to precessing BB Hs. We have performed aligned-spin searches using the SGNL pipeline on dat a containing simulated aligned-spin and precessing BBH signals\, producing matched-filter triggers for both classes. We propose to train a feed-forw ard neural network on the recovered trigger parameters\, such as the signa l-to-noise ratio\, consistency between the data and the best-fitting templ ate\, and recovered masses and spins\, to identify triggers likely to aris e from precessing signals. This classifier will serve as a preprocessing s tep to rapidly identify the small subset of triggers that require further follow-up. For this subset\, we propose an optimization method that combin es the signal-to-noise ratio with the autocorrelation-based least-squares signal consistency test to improve the ranking of precessing candidates.\n

We next address the characterization of transient bursts of detecto r noise\, or “glitches.” The LIGO detectors record data from more than 200\,000 auxiliary channels that monitor the instrument and its environme nt. We have developed a GPU-accelerated framework to transform time-series data from these auxiliary channels into the fractal dimension\, a compact representation of signal complexity. These representations will be used t o train an unsupervised autoencoder to learn nominal detector behavior and flag anomalous activity. We will use the detected anomalies to identify a uxiliary channels that witness distinct glitch populations\, pointing to t heir environmental or instrumental origin.

\n

Finally\, we address th e need for low-latency gravitational-wave alerts for multimessenger astron omy. Public alerts are currently issued approximately 30 seconds after a c andidate is identified using GWCelery\, a Python framework built on the Ce lery distributed task queue\, in which each stage records its result in th e GraceDB event database before the next proceeds. As part of the LIGO Sci entific Collaboration\, we are developing SGN-LLAI\, a streaming infrastru cture built on the Stream Graph Navigator (SGN) library\, in which results pass directly between stages via Kafka topics and are asynchronously reco rded in GraceDB\, removing database operations from the critical path. SGN -LLAI currently produces preliminary alerts in the standard public format\ , with latencies as low as 5 seconds in initial tests. The proposed next p hase will complete feature parity with GWCelery by adding source classific ation and properties\, data quality information\, and the full progression of public alerts from preliminary to initial and update notices\, includi ng human vetting and retraction. Alerts will then be disseminated through the GCN and SCiMMA networks.

\n

Advisor(s): Dr. Sarah Caudill\, Depar tment of Physics (scaudill@umassd.edu)                                                     

\n

Committee members:

\n\n

Note: All EAS Students are encouraged to attend.

Event page: https://www.umassd.edu/events/cms/10-22-26-eas-doctoral-proposal-d efense-by-anushka-doke.php

DTSTAMP:20261008T213030 DTSTART;TZID=America/New_York:20261022T120000 DTEND;TZID=America/New_York:20261022T140000 LOCATION:TXT 105 SUMMARY;LANGUAGE=en-us:EAS Doctoral Proposal Defense by Anushka Doke UID:7fdfe021d6c8e7e652402ae8e8eb5958@www.umassd.edu END:VEVENT END:VCALENDAR