BEGIN:VCALENDAR
VERSION:2.0
CALSCALE:GREGORIAN
PRODID:UW-Madison-Physics-Events
BEGIN:VEVENT
SEQUENCE:0
UID:UW-Physics-Event-8244
DTSTART:20230411T210000Z
DURATION:PT1H0M0S
DTSTAMP:20260414T053037Z
LAST-MODIFIED:20230404T180909Z
LOCATION:4274 Chamberlin
SUMMARY:A(ction) functional gradient descent algorithm for estimating 
 instantons in chemical reaction networks\, Graduate Program Event\, Pr
 aful Gagrani\, Physics Graduate Student
DESCRIPTION:Chemical reaction networks (CRNs) taken under mass-action 
 kinetics play a central role in the mathematical modeling of chemistry
  and biology. A key reason for their widespread utility is their capac
 ity to exhibit multiple attractors and capture a wide range of nonline
 ar phenomena. Computing paths of transition between attractors\, or in
 stantons\, is a challenging task\, not solvable analytically for all b
 ut the simplest cases. In our work\, we propose an algorithm for numer
 ically estimating instantons for a CRN. The algorithm uses the Hamilto
 nian description of a stochastic CRN and solves a MinMax problem on th
 e Action functional to converge on the instanton. In this talk\, I wil
 l present a schematic derivation of the Hamiltonian and Action functio
 nal for stochastic CRNs\, explain our Action Functional Gradient Desce
 nt (AFGD) algorithm\, and show computational and practical application
 s. Finally\, I will briefly discuss the unified formalism to which bot
 h stochastic and quantum Hamiltonians belong and propose directions fo
 r future research. (For details\, see https://journals.aps.org/pre/abs
 tract/10.1103/PhysRevE.107.034305.)
URL:https://www.physics.wisc.edu/events/?id=8244
END:VEVENT
END:VCALENDAR