[seminarioDeProbabilidad] Fwd: Seminario de Probabilidad y Procesos Estocásticos

[Maria Clara Fittipaldi]

---------- Forwarded message ---------
From: Maria Clara Fittipaldi <clarisssss en gmail.com>
Date: vie., 8 de marzo de 2019 7:42 p. m.
Subject: Seminario de Probabilidad y Procesos Estocásticos
To: <seminarioprobabilidadyprocesos en matem.unam.mx>



Miércoles 13 de Marzo a las 13h15, Salón S-104
Departamento de Matemáticas
Facultad de Ciencias

Tatiana González Grandon
Berlin Mathematical School

Nos hablará sobre "Dynamic Joint Probabilistic Constraint Optimization for
Hydro Reservoir Management"

Abstract:
A dynamic joint probabilistic constraint is an inequality of the type
P ( g i ( x 1 , x 2 ( ξ 1 ) x 2 ( ξ 1 , ξ 2 ) , … , x T ( ξ 1 , … , ξ T - 1
) , ξ 1 , … , ξ T ) ≤ 0 , i = 1 , … , m ) ≥ 0 ,
where ( ξ 1 , … , ξ T ) is a finite stochastic process, ( x 1 , … , x T ) is
an adapted process of decision policies depending on previously observed
outcomes of the random process, P is a probability measure and p ∈ [ 0,1 ] is
a probability level. A typical example arises in hydro power reservoir
control subject to level constraints where the above display figures as a
constraint in some optimization problem. The talk presents some structural
results for the associated probability function assigning to each set of
decision policies the probability occurring above. For instance, strong and
weak semicontinuity results are provided for the general case depending on
whether policies are supposed in L p or W 1 , p spaces. For a simple
two-stage model corresponding to the one of reservoir control, verifiable
conditions for Lipschitz continuity and differentiability of this
probability function are derived and endowed with explicit derivative
formulae. Numerical results are illustrated for the solution of such
two-stage problem.

Organizan
María Clara Fittipaldi
Yuri Salazar
Arno Siri-Jégousse
Geronimo Uribe Bravo


Página web:
http://www.matem.unam.mx/~seminarioproba/

Para suscribirse a la lista de distribución entrar a:
http://higgs.matem.unam.mx/cgi-bin/mailman/listinfo/seminarioprobabilidadyprocesos

y seguir el procedimiento.
------------ próxima parte ------------
Se ha borrado un adjunto en formato HTML...
URL: <http://higgs.matem.unam.mx/pipermail/seminarioprobabilidadyprocesos/attachments/20190310/16dea24d/attachment.html>