Learning to augment with feature side-information

Mollaysa, Amina (Haute école de gestion de Genève, HES-SO // Haute Ecole Spécialisée de Suisse Occidentale ; University of Geneva, Switzerland) ; Kalousis, Alexandros (Haute école de gestion de Genève, HES-SO // Haute Ecole Spécialisée de Suisse Occidentale ; University of Geneva, Switzerland) ; Bruno, Eric (Expedia, Switzerland) ; Diephuis, Maurits (University of Geneva, Switzerland)

Neural networks typically need huge amounts of data to train in order to get reasonable generalizable results. A common approach is to artificially generate samples by using prior knowledge of the data properties or other relevant domain knowledge. However, if the assumptions on the data properties are not accurate or the domain knowledge is irrelevant to the task at hand, one may end up degenerating learning performance by using such augmented data in comparison to simply training on the limited available dataset. We propose a critical data augmentation method using feature side-information, which is obtained from domain knowledge and provides detailed information about features' intrinsic properties. Most importantly, we introduce an instance wise quality checking procedure on the augmented data. It filters out irrelevant or harmful augmented data prior to entering the model. We validated this approach on both synthetic and real-world datasets, specifically in a scenario where the data augmentation is done based on a task independent, unreliable source of information. The experiments show that the introduced critical data augmentation scheme helps avoid performance degeneration resulting from incorporating wrong augmented data.


Conference Type:
full paper
Faculty:
Economie et Services
School:
HEG - Genève
Institute:
CRAG - Centre de Recherche Appliquée en Gestion
Subject(s):
Informatique
Publisher:
Nagoya, Japan, 17-19 November 2019
Date:
2019-11
Nagoya, Japan
17-19 November 2019
Pagination:
Pp. 173-187
Published in:
Proceedings of The Eleventh Asian Conference on Machine Learning
Numeration (vol. no.):
101
ISSN:
2640-3498
External resources:
Appears in Collection:



 Record created 2020-07-31, last modified 2020-10-27

Fulltext:
Download fulltext
PDF

Rate this document:

Rate this document:
1
2
3
 
(Not yet reviewed)