Deep Metric Learning-Based Semi-Supervised Regression with Alternate Learning
This paper introduces a novel deep metric learning-based semi-supervised regression (DML-S2R) method for parameter estimation problems. The proposed DML-S2R method aims to mitigate the problems of insufficient amount of labeled samples without collecting any additional sample with a target value. To...
Main Authors: | , , |
---|---|
Format: | Conference Object |
Language: | English |
Published: |
IEEE
2023
|
Subjects: | |
Online Access: | https://depositonce.tu-berlin.de/handle/11303/20635 https://doi.org/10.14279/depositonce-19433 |
id |
ftdepositonce:oai:depositonce.tu-berlin.de:11303/20635 |
---|---|
record_format |
openpolar |
spelling |
ftdepositonce:oai:depositonce.tu-berlin.de:11303/20635 2023-12-17T10:29:26+01:00 Deep Metric Learning-Based Semi-Supervised Regression with Alternate Learning Zell, Adina Sumbul, Gencer Demir, Begum 2023-11-15T10:08:09Z application/pdf https://depositonce.tu-berlin.de/handle/11303/20635 https://doi.org/10.14279/depositonce-19433 en eng IEEE 1522-4880 https://depositonce.tu-berlin.de/handle/11303/20635 https://doi.org/10.14279/depositonce-19433 2381-8549 http://rightsstatements.org/vocab/InC/1.0/ 000 Informatik Informationswissenschaft allgemeine Werke::000 Informatik Wissen Systeme::004 Datenverarbeitung Informatik semi-supervised regression parameter estimation metric learning deep learning Conference Object acceptedVersion 2023 ftdepositonce https://doi.org/10.14279/depositonce-19433 2023-11-20T17:19:31Z This paper introduces a novel deep metric learning-based semi-supervised regression (DML-S2R) method for parameter estimation problems. The proposed DML-S2R method aims to mitigate the problems of insufficient amount of labeled samples without collecting any additional sample with a target value. To this end, it is made up of two main steps: i) pairwise similarity modeling with scarce labeled data; and ii) triplet-based metric learning with abundant unlabeled data. The first step aims to model pairwise sample similarities by using a small number of labeled samples. This is achieved by estimating the target value differences of labeled samples with a Siamese neural network (SNN). The second step aims to learn a triplet-based metric space (in which similar samples are close to each other and dissimilar samples are far apart from each other) when the number of labeled samples is insufficient. This is achieved by employing the SNN of the first step for triplet-based deep metric learning that exploits not only labeled samples but also unlabeled samples. For the end-to-end training of DML-S2R, we investigate an alternate learning strategy for the two steps. Due to this strategy, the encoded information in each step becomes a guidance for learning phase of the other step. The experimental results confirm the success of DML-S2R compared to the state-of-the-art semi-supervised regression methods. The code of the proposed method is publicly available at https://git.tu-berlin.de/rsim/DML-S2R. EC/H2020/759764/EU/Accurate and Scalable Processing of Big Data in Earth Observation/BigEarth Conference Object DML TU Berlin: Deposit Once |
institution |
Open Polar |
collection |
TU Berlin: Deposit Once |
op_collection_id |
ftdepositonce |
language |
English |
topic |
000 Informatik Informationswissenschaft allgemeine Werke::000 Informatik Wissen Systeme::004 Datenverarbeitung Informatik semi-supervised regression parameter estimation metric learning deep learning |
spellingShingle |
000 Informatik Informationswissenschaft allgemeine Werke::000 Informatik Wissen Systeme::004 Datenverarbeitung Informatik semi-supervised regression parameter estimation metric learning deep learning Zell, Adina Sumbul, Gencer Demir, Begum Deep Metric Learning-Based Semi-Supervised Regression with Alternate Learning |
topic_facet |
000 Informatik Informationswissenschaft allgemeine Werke::000 Informatik Wissen Systeme::004 Datenverarbeitung Informatik semi-supervised regression parameter estimation metric learning deep learning |
description |
This paper introduces a novel deep metric learning-based semi-supervised regression (DML-S2R) method for parameter estimation problems. The proposed DML-S2R method aims to mitigate the problems of insufficient amount of labeled samples without collecting any additional sample with a target value. To this end, it is made up of two main steps: i) pairwise similarity modeling with scarce labeled data; and ii) triplet-based metric learning with abundant unlabeled data. The first step aims to model pairwise sample similarities by using a small number of labeled samples. This is achieved by estimating the target value differences of labeled samples with a Siamese neural network (SNN). The second step aims to learn a triplet-based metric space (in which similar samples are close to each other and dissimilar samples are far apart from each other) when the number of labeled samples is insufficient. This is achieved by employing the SNN of the first step for triplet-based deep metric learning that exploits not only labeled samples but also unlabeled samples. For the end-to-end training of DML-S2R, we investigate an alternate learning strategy for the two steps. Due to this strategy, the encoded information in each step becomes a guidance for learning phase of the other step. The experimental results confirm the success of DML-S2R compared to the state-of-the-art semi-supervised regression methods. The code of the proposed method is publicly available at https://git.tu-berlin.de/rsim/DML-S2R. EC/H2020/759764/EU/Accurate and Scalable Processing of Big Data in Earth Observation/BigEarth |
format |
Conference Object |
author |
Zell, Adina Sumbul, Gencer Demir, Begum |
author_facet |
Zell, Adina Sumbul, Gencer Demir, Begum |
author_sort |
Zell, Adina |
title |
Deep Metric Learning-Based Semi-Supervised Regression with Alternate Learning |
title_short |
Deep Metric Learning-Based Semi-Supervised Regression with Alternate Learning |
title_full |
Deep Metric Learning-Based Semi-Supervised Regression with Alternate Learning |
title_fullStr |
Deep Metric Learning-Based Semi-Supervised Regression with Alternate Learning |
title_full_unstemmed |
Deep Metric Learning-Based Semi-Supervised Regression with Alternate Learning |
title_sort |
deep metric learning-based semi-supervised regression with alternate learning |
publisher |
IEEE |
publishDate |
2023 |
url |
https://depositonce.tu-berlin.de/handle/11303/20635 https://doi.org/10.14279/depositonce-19433 |
genre |
DML |
genre_facet |
DML |
op_relation |
1522-4880 https://depositonce.tu-berlin.de/handle/11303/20635 https://doi.org/10.14279/depositonce-19433 2381-8549 |
op_rights |
http://rightsstatements.org/vocab/InC/1.0/ |
op_doi |
https://doi.org/10.14279/depositonce-19433 |
_version_ |
1785581808121806848 |