Skip to Main content Skip to Navigation
Reports

Unsupervised threshold determination for hybrid models

Abstract : A hybrid Gauss-Pareto model is considered for asymmetric heavy tailed data. The paper presents an unsupervised itera- tive algorithm to find successively the parameters of the Gaus- sian density and that of the Generalized Pareto distribution (GPD) with a continuity constrain on the hybrid density and its derivative at the junction point. Simulation results show that the proposed iterative algorithm provides reliable posi- tion for the junction point as well as an accurate estimation of the GPD parameters, compared to state of the art methods. Furthermore, a great advantage of the proposed method is that it can be adapted to any hybrid model.
Document type :
Reports
Complete list of metadata

Cited literature [21 references]  Display  Hide  Download

https://hal.inria.fr/hal-01367584
Contributor : Mamadou Mboup <>
Submitted on : Friday, September 16, 2016 - 1:59:09 PM
Last modification on : Monday, May 17, 2021 - 2:52:11 PM
Long-term archiving on: : Saturday, December 17, 2016 - 3:03:29 PM

File

unsupervised_hybrid_2013.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01367584, version 1

Collections

Citation

Nehla Debbabi, Marie Kratz, Mamadou Mboup. Unsupervised threshold determination for hybrid models. [Research Report] Université de Reims Champagne Ardenne URCA. 2013. ⟨hal-01367584⟩

Share

Metrics

Record views

155

Files downloads

119