Publications
2013
1.
González-Castro, Víctor; Alaiz-Rodríguez, Rocío; Alegre, Enrique
Class distribution estimation based on the Hellinger distance Artículo de revista
En: Information Sciences, vol. 218, pp. 146–164, 2013, (Publisher: Elsevier).
Resumen | Enlaces | BibTeX | Etiquetas: Class Distribution Shift, Class Prior Probability Estimation, Hellinger Distance, Quantification
@article{gonzalez-castro_class_2013,
title = {Class distribution estimation based on the Hellinger distance},
author = {Víctor González-Castro and Rocío Alaiz-Rodríguez and Enrique Alegre},
url = {https://www.sciencedirect.com/science/article/pii/S0020025512004069},
year = {2013},
date = {2013-01-01},
journal = {Information Sciences},
volume = {218},
pages = {146–164},
abstract = {This work presents a novel quantification method for two-class problems using distributional divergence measures. It estimates class priors by minimizing Hellinger distance between test and controlled validation distributions. Experiments, including a boar semen quality control task, show improved accuracy over existing methods.},
note = {Publisher: Elsevier},
keywords = {Class Distribution Shift, Class Prior Probability Estimation, Hellinger Distance, Quantification},
pubstate = {published},
tppubtype = {article}
}
This work presents a novel quantification method for two-class problems using distributional divergence measures. It estimates class priors by minimizing Hellinger distance between test and controlled validation distributions. Experiments, including a boar semen quality control task, show improved accuracy over existing methods.
2011
2.
Alegre, Enrique
Descripción adaptativa de texturas y estimación de las probabilidades a priori de las clases para el control de calidad seminal Tesis doctoral
Universidad de León, 2011.
Resumen | Enlaces | BibTeX | Etiquetas: Calidad Seminal, Descripción Adaptativa de Texturas, Estimación, Image Texture, Quantification, segmentation, Unlabeled Datasets
@phdthesis{alegre_descripcion_2011,
title = {Descripción adaptativa de texturas y estimación de las probabilidades a priori de las clases para el control de calidad seminal},
author = {Enrique Alegre},
url = {https://scholar.google.es/citations?view_op=view_citation&hl=en&user=81rvBFwAAAAJ&cstart=20&pagesize=80&sortby=title&citation_for_view=81rvBFwAAAAJ:IjCSPb-OGe4C},
year = {2011},
date = {2011-01-01},
school = {Universidad de León},
abstract = {In this Thesis we have evaluated several approaches to describe digital image textures. In addition, we have proposed a new intelligent segmentation procedure, an original adaptive texture descriptor and two new methods for estimating class proportions (quantification) of unlabelled datasets.},
keywords = {Calidad Seminal, Descripción Adaptativa de Texturas, Estimación, Image Texture, Quantification, segmentation, Unlabeled Datasets},
pubstate = {published},
tppubtype = {phdthesis}
}
In this Thesis we have evaluated several approaches to describe digital image textures. In addition, we have proposed a new intelligent segmentation procedure, an original adaptive texture descriptor and two new methods for estimating class proportions (quantification) of unlabelled datasets.