The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
La représentation clairsemée conjointe multitâche (MTJSR) est un type de méthode d'apprentissage multitâche (MTL) efficace pour résoudre différents problèmes ensemble à l'aide d'une représentation clairsemée partagée. Basé sur le mécanisme d'apprentissage chez l'humain, qui est un apprentissage à votre rythme en entraînant progressivement les tâches de facile à difficile, j'applique ce mécanisme dans MTJSR et propose une représentation clairsemée conjointe multi-tâches avec un apprentissage à votre rythme (MTJSR-SP ) algorithme. Dans MTJSR-SP, le mécanisme d'apprentissage à votre rythme est considéré comme un régularisateur de la fonction d'optimisation, et une optimisation itérative est appliquée pour le résoudre. Par rapport aux méthodes MTL traditionnelles, MTJSR-SP est plus robuste au bruit et aux valeurs aberrantes. Les résultats expérimentaux sur certains ensembles de données, à savoir deux ensembles de données synthétisés, quatre ensembles de données du référentiel d'apprentissage automatique UCI, un ensemble de données de fleurs d'Oxford et un ensemble de données de catégorisation d'images Caltech-256, sont utilisés pour valider l'efficacité de MTJSR-SP.
Lihua GUO
South China University of Technology
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copier
Lihua GUO, "From Easy to Difficult: A Self-Paced Multi-Task Joint Sparse Representation Method" in IEICE TRANSACTIONS on Information,
vol. E101-D, no. 8, pp. 2115-2122, August 2018, doi: 10.1587/transinf.2017EDP7289.
Abstract: Multi-task joint sparse representation (MTJSR) is one kind of efficient multi-task learning (MTL) method for solving different problems together using a shared sparse representation. Based on the learning mechanism in human, which is a self-paced learning by gradually training the tasks from easy to difficult, I apply this mechanism into MTJSR, and propose a multi-task joint sparse representation with self-paced learning (MTJSR-SP) algorithm. In MTJSR-SP, the self-paced learning mechanism is considered as a regularizer of optimization function, and an iterative optimization is applied to solve it. Comparing with the traditional MTL methods, MTJSR-SP has more robustness to the noise and outliers. The experimental results on some datasets, i.e. two synthesized datasets, four datasets from UCI machine learning repository, an oxford flower dataset and a Caltech-256 image categorization dataset, are used to validate the efficiency of MTJSR-SP.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2017EDP7289/_p
Copier
@ARTICLE{e101-d_8_2115,
author={Lihua GUO, },
journal={IEICE TRANSACTIONS on Information},
title={From Easy to Difficult: A Self-Paced Multi-Task Joint Sparse Representation Method},
year={2018},
volume={E101-D},
number={8},
pages={2115-2122},
abstract={Multi-task joint sparse representation (MTJSR) is one kind of efficient multi-task learning (MTL) method for solving different problems together using a shared sparse representation. Based on the learning mechanism in human, which is a self-paced learning by gradually training the tasks from easy to difficult, I apply this mechanism into MTJSR, and propose a multi-task joint sparse representation with self-paced learning (MTJSR-SP) algorithm. In MTJSR-SP, the self-paced learning mechanism is considered as a regularizer of optimization function, and an iterative optimization is applied to solve it. Comparing with the traditional MTL methods, MTJSR-SP has more robustness to the noise and outliers. The experimental results on some datasets, i.e. two synthesized datasets, four datasets from UCI machine learning repository, an oxford flower dataset and a Caltech-256 image categorization dataset, are used to validate the efficiency of MTJSR-SP.},
keywords={},
doi={10.1587/transinf.2017EDP7289},
ISSN={1745-1361},
month={August},}
Copier
TY - JOUR
TI - From Easy to Difficult: A Self-Paced Multi-Task Joint Sparse Representation Method
T2 - IEICE TRANSACTIONS on Information
SP - 2115
EP - 2122
AU - Lihua GUO
PY - 2018
DO - 10.1587/transinf.2017EDP7289
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E101-D
IS - 8
JA - IEICE TRANSACTIONS on Information
Y1 - August 2018
AB - Multi-task joint sparse representation (MTJSR) is one kind of efficient multi-task learning (MTL) method for solving different problems together using a shared sparse representation. Based on the learning mechanism in human, which is a self-paced learning by gradually training the tasks from easy to difficult, I apply this mechanism into MTJSR, and propose a multi-task joint sparse representation with self-paced learning (MTJSR-SP) algorithm. In MTJSR-SP, the self-paced learning mechanism is considered as a regularizer of optimization function, and an iterative optimization is applied to solve it. Comparing with the traditional MTL methods, MTJSR-SP has more robustness to the noise and outliers. The experimental results on some datasets, i.e. two synthesized datasets, four datasets from UCI machine learning repository, an oxford flower dataset and a Caltech-256 image categorization dataset, are used to validate the efficiency of MTJSR-SP.
ER -