Descripción
|
|
---|---|
By inference in BNs we refer to the task of computing a-posteriori probabilities. This task can be found under different names: probability propagation, belief updating, belief revision, ... Most of queries involve observations or evidence Evidence on a variable is a statement of the certainties of its states, i.e., (flu=yes). Hard evidence. An evidence function that assigns a zero probability to all but one state is said to provide hard evidence. Hard evidence e over a set of variables E is often referred to as instantiation. temperature=high, headache=no Soft evidence. An evidence function that assigns a probability distribution over dom(Ei) for each Ei ¿ E is said to provide soft evidence. - Example. If dom(temperature) = (no, ligth, high, very-high) and dom(headache)=(yes,no), the we can have the following soft evidence: - se(temperature) = (0,0,2,1) and se(headache) = (2,1) | |
Internacional
|
Si |
Nombre congreso
|
Advanced Data Analysis and Modelling Summerschool |
Entidad organizadora
|
Facultad de Informática, Universidad Politécnica de Madrid |
Nacionalidad Entidad
|
ESPAÑA |
Lugar/Ciudad de impartición
|
Madrid |
Fecha inicio
|
05/07/2009 |
Fecha fin
|
09/07/2009 |