BEGIN:VCALENDAR VERSION:2.0 PRODID:-//linuxsoftware.nz//NONSGML Joyous v1.4//EN BEGIN:VEVENT SUMMARY:Fink broker\, an optimized recommendation system for transient fol low-up DTSTART:20220422T143000Z DTEND:20220422T163000Z DTSTAMP:20260801T131108Z UID:0100e939-95e6-4d28-a1e8-e30f17e5653b SEQUENCE:1 CREATED:20220422T091324Z DESCRIPTION: ABSTRACT: Next generation experiments such as the Vera Rubin Observatory Legacy Survey of Space and Time (LSST) will provide an unprece dented volume of time-domain data opening a new era of big data in astrono my. To fully harness the power of these surveys\, we require analysis meth ods capable of dealing with large data volumes that can identify promising transients within minutes for follow-up coordination. In this talk I will describe the infrastructure put in place by LSST to handle approximately 10 million transient candidates per night and present Fink\, a broker deve loped to face these challenges. Fink is based on high-end technology and d esigned for fast and efficient analysis of big flows. It has been chosen a s one of the official LSST brokers and will receive the full data stream. I will highlight the state-of-the-art machine learning techniques used to generate early classification scores for a variety of time-domain phenomen a including kilonovae and supernovae\, as well as for artifacts\, like sat ellites glitches. Such methods include Deep Learning advances and Active L earning approaches to coherently incorporate available information\, deliv ering increasingly more accurate added values throughout the duration of t he survey. I will also highlight the potential for discovery of new catego ries of sources and how we can optimize for discovery in the era of LSST. LAST-MODIFIED:20220422T091324Z LOCATION:Sala de Seminários do DF\, Pavilhão de Física\, 2º piso URL:http://df.vps.tecnico.ulisboa.pt/pt/eventos/fink-broker-an-optimized-r ecommendation-system-for-transient-follow-up/ X-ALT-DESC;FMTTYPE=text/html:

ABSTRACT: Next gen eration experiments such as the Vera Rubin Observatory Legacy Survey of Sp ace and Time (LSST) will provide an unprecedented volume of time-domain da ta opening a new era of big data in astronomy. To fully harness the power of these surveys\, we require analysis methods capable of dealing with lar ge data volumes that can identify promising transients within minutes for follow-up coordination. In this talk I will describe the infrastructure pu t in place by LSST to handle approximately 10 million transient candidates per night and present Fink\, a broker developed to face these challenges. Fink is based on high-end technology and designed for fast and efficient analysis of big flows. It has been chosen as one of the official LSST brok ers and will receive the full data stream. I will highlight the state-of-t he-art machine learning techniques used to generate early classification s cores for a variety of time-domain phenomena including kilonovae and super novae\, as well as for artifacts\, like satellites glitches. Such methods include Deep Learning advances and Active Learning approaches to coherentl y incorporate available information\, delivering increasingly more accurat e added values throughout the duration of the survey. I will also highligh t the potential for discovery of new categories of sources and how we can optimize for discovery in the era of LSST.

END:VEVENT END:VCALENDAR