BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:Identifying statistical indicators of temporal asymmetry using a d
 ata-driven approach
DTSTART:20261019T123000Z
DTEND:20261019T133000Z
DTSTAMP:20261023T052700Z
UID:indico-event-1950@indico.dfa.unipd.it
CONTACT:marco.baiesi@unipd.it
DESCRIPTION:Speakers: Teresa Dalle Nogare (School of Physics\, The Univers
 ity of Sydney)\n\nThe dynamics of time-reversible systems are statisticall
 y indistinguishable when observed forward or backward in time. A rich lite
 rature of statistical methods to distinguish irreversible dynamics from th
 e reversible dynamics of linear Gaussian processes can provide insights in
 to underlying mechanisms and aid modeling and statistical quantification o
 f time-series data. However\, existing time-reversibility metrics have bee
 n developed individually\, forming a fragmented body of research that make
 s it challenging to identify the most effective approaches developed to da
 te\, and the most promising directions for development. Here\, we address 
 these issues by systematically evaluating over 6000 time-series summary st
 atistics\, derived from across the time-series analysis literature\, on th
 eir ability to distinguish the time irreversibility of data simulated from
  a diverse range of 35 systems. Our large-scale data-driven comparison hig
 hlights the effectiveness of several key families of statistics\, includin
 g time-asymmetric forms of generalized autocorrelation functions\, time-se
 ries symbolic sequences\, and forecasting-related methods. All irreversibl
 e systems studied here could be accurately distinguished by a well-chosen 
 time-series statistic\, but no single statistic could accurately index the
  statistical form of irreversibility for all irreversible systems. This ch
 allenges the assumption that a given time-reversibility statistic will acc
 urately capture time reversibility in general\, and underscores the import
 ance of tailoring statistical approaches to the time-reversal characterist
 ics of a given system. Our results provide a unified understanding of the 
 key algorithmic structures through which irreversibility can be effectivel
 y quantified from data\, providing a foundation for connecting patterns in
  time series to the underlying mechanisms of the systems that generate the
 m.\n\nhttps://indico.dfa.unipd.it/event/1950/
LOCATION:P1D (Dipartimento di Fisica e Astronomia - Edificio Ricci-Curbast
 ro (ex-Paolotti))
URL:https://indico.dfa.unipd.it/event/1950/
END:VEVENT
END:VCALENDAR