Mixture of Gaussians for Distance Estimation ...
Document type :
Compte-rendu et recension critique d'ouvrage
Title :
Mixture of Gaussians for Distance Estimation with Missing Data
Author(s) :
Eirola, Emil [Auteur correspondant]
School of Electrical Engineering [Aalto Univ]
Lendasse, Amaury [Auteur]
Laboratory of Computer and Information Science [CIS]
Vandewalle, Vincent [Auteur]
MOdel for Data Analysis and Learning [MODAL]
Université de Lille, Droit et Santé
Biernacki, Christophe [Auteur]
MOdel for Data Analysis and Learning [MODAL]
School of Electrical Engineering [Aalto Univ]
Lendasse, Amaury [Auteur]
Laboratory of Computer and Information Science [CIS]
Vandewalle, Vincent [Auteur]
MOdel for Data Analysis and Learning [MODAL]
Université de Lille, Droit et Santé
Biernacki, Christophe [Auteur]
MOdel for Data Analysis and Learning [MODAL]
Journal title :
Neurocomputing
Pages :
32-42
Publisher :
Elsevier
Publication date :
2014-05-05
ISSN :
0925-2312
English keyword(s) :
Missing data
distance estimation
mixture model
distance estimation
mixture model
HAL domain(s) :
Statistiques [stat]/Méthodologie [stat.ME]
English abstract : [en]
The majority of all commonly used machine learning methods can not be applied directly to data sets with missing values. However, most such meth- ods only depend on the relative di erences between samples instead of their ...
Show more >The majority of all commonly used machine learning methods can not be applied directly to data sets with missing values. However, most such meth- ods only depend on the relative di erences between samples instead of their particular values, and thus one useful approach is to directly estimate the pairwise distances between all samples in the data set. This is accomplished by tting a Gaussian mixture model to the data, and using it to derive estimates for the distances. Experimental simulations con rm that the pro- posed method provides accurate estimates compared to alternative methods for estimating distances. The experimental evaluation additionally shows that more accurately estimating distances leads to improved prediction performance for classification and regression tasks when used as inputs for a neural network.Show less >
Show more >The majority of all commonly used machine learning methods can not be applied directly to data sets with missing values. However, most such meth- ods only depend on the relative di erences between samples instead of their particular values, and thus one useful approach is to directly estimate the pairwise distances between all samples in the data set. This is accomplished by tting a Gaussian mixture model to the data, and using it to derive estimates for the distances. Experimental simulations con rm that the pro- posed method provides accurate estimates compared to alternative methods for estimating distances. The experimental evaluation additionally shows that more accurately estimating distances leads to improved prediction performance for classification and regression tasks when used as inputs for a neural network.Show less >
Language :
Anglais
Popular science :
Non
Collections :
Source :
Files
- document
- Open access
- Access the document
- Publication206.pdf
- Open access
- Access the document
- mdmixmod-neuroc.pdf
- Open access
- Access the document