<html>
<head>
<meta http-equiv="content-type" content="text/html; charset=UTF-8">
</head>
<body bgcolor="#FFFFFF" text="#000000">
Estimados Académicos y alumnos,<br>
<br>
<br>
Se les recuerda que hoy Jueves 03 de Noviembre del 2016 a las
14:00 hrs, se realizará la charla del Grupo de lectura aprendizaje
de máquinas, en la Sala Multimedia del Sexto Piso, Beauchef 851,
Torre Norte.<br>
<p><strong><u>GLAM #7: </u> <br>
</strong></p>
<p><strong>Sequential Monte Carlo Methods: Particle Filtering
(Joaquín Rojas)</strong></p>
<p><strong></strong><strong></strong><strong>Fecha: </strong>3/11/16</p>
<p><strong>Lugar:</strong> Sala multimedia piso 6, CMM</p>
<p><strong>hora: </strong>1400</p>
<p> </p>
<br>
<p><b>Abstract:</b> We review the main sequential Monte Carlo
methods existing in the literature. In particular, we are
interested in Bayesian filtering of state-space models, with
potentially nonlinear measurement and transition functions, as
well as non Gaussian innovations. We first review the state-space
framework and the filtering problem. We then introduce importance
sampling based methods for sequential updating of the filtering
distribution, such as the original bootstrap filter of Gordon et.
al. (1993), and the auxiliary particle filter of Pitt and Shephard
(1999). We discuss particle degeneracy and common problems
associated with particle filter implementation. Finally, we
conclude with a brief overview of recent contributions.<br>
</p>
<br>
<p><b>Referencias: </b><br>
</p>
<p>[1] Gordon, Salmond and Smith (1993): "A Novel Approach to
Nonlinear and Non-Gaussian Bayesian State Estimation", IEE
Procedings<br>
</p>
[2] Pitt and Shephard (1999): "Filtering Via Simulation: Auxiliary
Particle Filters", Journal of the American Statistical Association<br>
<br>
Esperando contar con su presencia, les saluda, <br>
<br>
Ma. Inés Rivera <br>
<br>
<br>
</body>
</html>