Seminario de Teoría Espectral


2016-05-05
Christian Sadel. Facultad de Matemáticas, PUC
Sde Limits for Products of Random Matrices and Goe Statistics for Rescaled Anderson Models on Long Strips.
Sala 1 de la Facultad de Matemáticas PUC a las 17 horas .
2016-04-28
Carlos Villegas-Blas. Universidad Nacional Autónoma de México
Sobre Teoremas de Distribución Límite de Autovalores Para el Hidrógeno en Campos Eléctricos O Magnéticos Constantes.
Sala 1 de la Facultad de Matemáticas a las 17:00 Hrs.
2015-12-03
Giuseppe de Nittis. Facultad de Matemáticas, PUC
Topological Piezo-Currents in Graphene
Sala 1 - Facultad de Matemáticas a las 17:00 Hrs.
2015-11-19
Marcus Carlsson. Universidad de Lund, Suecia
Multidimensional Frequency Estimation Using General Domain Hankel and Toeplitz Operators
Sala 1 de la Facultad de Matemáticas a las 17:00 Hrs.
2015-11-12
Werner Kirsch. Universidad de Hagen, Alemania
On The Mathematics of Political Power
Sala 1 de la Facultad de Matemáticas a las 17:00 Hrs.
2015-11-05
Rafael Tiedra de Aldecoa. Facultad de Matemáticas, PUC
Spectral Properties of Horocycle Flows for Surfaces of Negative Curvature
Sala 1 de la Facultad de Matemáticas a las 17:00 Hrs.
2015-10-29
Erdal Emsiz. Facultad de Matemáticas, PUC
Difference Equation for The Heckman-Opdam Hypergeometric Function and its Confluent Whittaker Limit
Sala 1 de la Facultad de Matemáticas a las 17:00 Hrs.
2015-10-22
Søren Fournais. Aarhus University
The Semi-Classical Limit of Large Fermionic Systems
Sala 1 de la Facultad de Matemáticas a las 17:00 Hrs.
2015-10-15
Rafael Benguria. Instituto de Fisica, PUC
Energia del Estado Fundamental de Un Sistema de Polarones
Sala 1 de la Facultad de Matemáticas UC a las 17:00 Hrs
2015-08-27
Rafael Tiedra de Aldecoa. P. Universidad Católica de Chile
Spectral and Scattering Properties At Thresholds for The Laplacian in a Half-Space With a Periodic Boundary Condition
Sala 1 de la Facultad de Matemáticas UC a las 17:00 Hrs.
2015-08-20
Rafael D. Benguria. Instituto de Física, PUC
The Brezis-Nirenberg Problem on S^n, in Spaces of Fractional Dimension
Sala 1 Facultad de Matemáticas UC - 17:00 Hrs.
2015-06-25
Marius Mantoiu. Universidad de Chile
On The Essential Spectrum.
Sala 1 a las 17:00 Hrs.
2015-04-30
Christian Sadel. Institute of Sience and Technology, Klosterneuburg, Austria
Anderson Transition At 2 Dimensional Growth Rate on Antitrees With Normalized Edge Weights
Sala 1 Facultad de Matemáticas 17:00 Hrs.
2014-10-09
Diomba Sambou. Facultad de Matemáticas, PUC
Lieb-Thirring Type Inequalities for Non Self-Adjoint Schrödinger Operators
Sala 1 Facultad de Matemáticas 17:00 Hrs.
2014-07-03
Manuel Larenas. Rutgers University.
Asymptotic Stability of Soliton States of Nonlinear Schrödinger Equations
Sala 1 Facultad de Matemáticas - 17:00 Hrs.
2014-05-29
Enrique Muñoz. Instituto de Física, PUC
Transporte Cuántico en Sistemas de Baja Dimensionalidad
Sala 1, Facultad de Matemáticas. - 17:00 Hrs.
2014-05-08
Yulia Karpeshina. University of Alabama At Birmingham, Usa
On Schroedinger Operator With Quasi-Periodic Potential in Dimension Two
Sala 1 - Facultad de Matemáticas - 17:00 Hrs.
2014-04-10
Rafael Tiedra de Aldecoa. Facultad de Matemáticas, PUC
Resolvent Expansions and Continuity of The Scattering Matrix At Embedded Thresholds
Sala 1 Facultad de Matemáticas - 17:00 hrs.
2014-03-27
Francisco Hoecker. Tu Chemnitz
Asintóticas de Lifschitz en la Red de Bethe
Sala 1 de la Facultad de Matemáticas - 17:00 hrs.
2013-12-19
Carlos Sing-Long. Stanford University
Risk estimation for regularized regression problems
Sala 1, Facultad de Matemáticas - 17:00 Hrs.
Abstract:
Abstract:
In many problems in science and engineering one wants to recover an object from incomplete information obtained from linear measurements. In practice the measurements are corrupted by noise and therefore exact recovery is not possible. When the underlying object has some a priori known structure, a popular approach is to use a regularized maximum likelihood estimator obtained by solving a convex optimization problem. The objective function consists of two terms, one that enforces data consistency, usually the likelihood, and another that enforces the known structure in the object. Typically this trade-off is controlled by a non-negative scalar multiplying the regularizer. This procedure yields a family of estimates parametrized by the value of this scalar. Intuitively,