BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Data driven mesoscale modelling of collective movement - Arshed Na
 beel (Indian Institute of Science)
DTSTART:20230808T135000Z
DTEND:20230808T141000Z
UID:TALK203122@talks.cam.ac.uk
DESCRIPTION:In the study of collective behaviour\, a central question is t
 o understand how group properties emerge from simple individual interactio
 ns at the level of the individuals. Usually\, one defines an order paramet
 er that quantifies some property of the group\, and is interested in study
 ing the dynamics of this order parameter. Often\, the group-level dynamica
 l equations are analytically derived from first principles\, based on assu
 mptions on the local interactions. However\, there are two main difficulti
 es to this approach. First\, analytical derivations are often based on man
 y simplifying assumptions on the individual interactions\, as all but the 
 simplest of models become analytically intractable. Second\, when studying
  collective behaviour in the real world (e.g. collective animal behaviour)
 \, the group sizes are typically small\, where the individual-level stocha
 sticity can have observable effects on the order parameter. Indeed\, in re
 al world systems\, the exact individual interactions are usually unknown\,
  adding another layer of difficulty.\nWe circumvent these difficulties wit
 h a novel data-driven approach to directly derive stochastic differential 
 equations (SDEs) from observed time series data. By combining techniques f
 rom machine learning with physical understanding of the system (such as in
 herent symmetries)\, we are able to derive accurate SDE models from observ
 ed or simulated time series data. I will present some of our recent work o
 n characterizing the mesoscale dynamics of both real-world and simulated f
 ish schools. In real fish schooling experiments\, we observe both unifying
  and differentiating aspects in the group dynamics of different species. T
 he determinstic dynamics of the group differs across species\, and can pro
 mote or destroy group-level order depending on species. On the other hand\
 , the stochastic dynamics typically have a characteristic structure that p
 romote order in small- to intermediate-sized groups. By connecting these o
 bservations to the mesoscale group dynamics of simulated fish schools\, we
  make putative inferences about how the inter-individual interactions diff
 er across different species.
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR