Skip to Main content Skip to Navigation
Conference papers

Sur la recherche de φ-entropie à maximisante donnée

Abstract : In this paper, we are interested in maximum entropy problems under moment constraints. Contrary to the usual problem of finding the maximizer of a given entropy, or of selecting constraints such that a given distribution is a maximizer, we focus here on the determination of an entropy such that a given distribution is its maximizer. The goal is in some sense to adapt the entropy to its maximizer, with potential application in entropy-based goodness-of-fit tests. It allows us to consider distributions out the exponential family – to which the maximizers of the Shannon entropy belong, and also to consider simple moment constraints, estimated from the observed sample. Finally, this approach also yields entropic functionals that are function of both probability density and state, allowing us to include skew-symmetric or multimodal distributions in the setting.
Complete list of metadatas

https://hal.archives-ouvertes.fr/hal-01221001
Contributor : Steeve Zozor <>
Submitted on : Tuesday, October 27, 2015 - 11:28:12 AM
Last modification on : Wednesday, May 13, 2020 - 4:30:05 PM

Identifiers

  • HAL Id : hal-01221001, version 1

Citation

Jean-François Bercher, Valerie Girardin, Justine Lequesne, Philippe Regnault, Steeve Zozor. Sur la recherche de φ-entropie à maximisante donnée. XXVème colloque GRETSI (GRETSI 2015), Sep 2015, Lyon, France. pp.n/c. ⟨hal-01221001⟩

Share

Metrics

Record views

525