ProbXiv
Explorersign in
Problem archiveProblem record

Statement

An open problem is to consider the limit ∆t \to 0 and derive a time-continuous description of the stochastic multi-agent system in Bures–Wasserstein space. We consider below a perturbed version of the update where we introduce a ran- i,Σ dom tangent vector W(k) \in Sym component-wise normally distributed. To ensure the arguments of the exponential map remains inside the definition domain {T \in Sym | T ≻ I} we introduce for \varepsilon > 0 the closed set Ω\varepsilon = {T \in Sym | T −Id + \varepsilonId } and the projection map ΠΩ\varepsilon (T ) := argminS\inΩ\varepsilon kT − SkΣ . The time-discrete con- sensus dynamics with additional stochasticity then reads for \sigma(k) > 0  1 1 N N   N m (k) , Σ (k) = Barycenter \omega N (m (k) , Σ ), . . . , N (m , Σ )    \sqrt{ }(k) i,m (k) (k)  mi = m i + ∆t(m (k) − m i ) + \sigma (k) ∆tW \forall i = 1, . ..,N (k+1) (k) (k) (k) (4) i BW \sqrt{ } i,Σ   T(k+1) = ΠΩ\varepsilon ∆t log (Σ(k) ) + \sigma(k) ∆tW(k) \forall i = 1, . . . , N     Σi BW i (k+1) = expΣi T(k+1) \forall i = 1, . . . , N . (k)

Record

Comments

Nothing has been published against this problem yet, and nothing has been said about it.

Sign in with an institutional address to take part in the discussion. Reading every thread stays open to everyone.

Sign in

Solve with an agent

Open the statement in a chat, with the problem and the ground rules already written into the prompt.

This opens a third-party site. Nothing is posted back to ProbXiv and nothing you write there is recorded here — what a model gives you is an attempt, which a person still has to check.