Details
| Original language | English |
|---|---|
| Article number | 111957 |
| Pages (from-to) | 111957 |
| Journal | Applied soft computing |
| Volume | 165 |
| Early online date | 8 Jul 2024 |
| Publication status | E-pub ahead of print - 8 Jul 2024 |
| Externally published | Yes |
Abstract
As a result of the ever increasing complexity of configuring and fine-tuning machine learning models, the field of automated machine learning (AutoML) has emerged over the past decade. However, software implementations like Auto-WEKA and Auto-sklearn typically focus on classical machine learning (ML) tasks such as classification and regression. Our work can be seen as the first attempt at offering a single AutoML framework for most problem settings that fall under the umbrella of multi-target prediction, which includes popular ML settings such as multi-label classification, multivariate regression, multi-task learning, dyadic prediction, matrix completion, and zero-shot learning. Automated problem selection and model configuration are achieved by extending DeepMTP, a general deep learning framework for MTP problem settings, with popular hyperparameter optimization (HPO) methods. Our extensive benchmarking across different datasets and MTP problem settings identifies cases where specific HPO methods outperform others.
Keywords
- Automated machine learning, Hyperparameter optimization, Machine learning, Multi-label classification, Multi-target prediction
ASJC Scopus subject areas
- Computer Science(all)
- Software
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
In: Applied soft computing, Vol. 165, 111957, 11.2024, p. 111957.
Research output: Contribution to journal › Article › Research › peer review
}
TY - JOUR
T1 - Hyperparameter optimization of two-branch neural networks in multi-target prediction
AU - Iliadis, Dimitrios
AU - Wever, Marcel
AU - De baets, Bernard
AU - Waegeman, Willem
N1 - Publisher Copyright: © 2024 Elsevier B.V.
PY - 2024/7/8
Y1 - 2024/7/8
N2 - As a result of the ever increasing complexity of configuring and fine-tuning machine learning models, the field of automated machine learning (AutoML) has emerged over the past decade. However, software implementations like Auto-WEKA and Auto-sklearn typically focus on classical machine learning (ML) tasks such as classification and regression. Our work can be seen as the first attempt at offering a single AutoML framework for most problem settings that fall under the umbrella of multi-target prediction, which includes popular ML settings such as multi-label classification, multivariate regression, multi-task learning, dyadic prediction, matrix completion, and zero-shot learning. Automated problem selection and model configuration are achieved by extending DeepMTP, a general deep learning framework for MTP problem settings, with popular hyperparameter optimization (HPO) methods. Our extensive benchmarking across different datasets and MTP problem settings identifies cases where specific HPO methods outperform others.
AB - As a result of the ever increasing complexity of configuring and fine-tuning machine learning models, the field of automated machine learning (AutoML) has emerged over the past decade. However, software implementations like Auto-WEKA and Auto-sklearn typically focus on classical machine learning (ML) tasks such as classification and regression. Our work can be seen as the first attempt at offering a single AutoML framework for most problem settings that fall under the umbrella of multi-target prediction, which includes popular ML settings such as multi-label classification, multivariate regression, multi-task learning, dyadic prediction, matrix completion, and zero-shot learning. Automated problem selection and model configuration are achieved by extending DeepMTP, a general deep learning framework for MTP problem settings, with popular hyperparameter optimization (HPO) methods. Our extensive benchmarking across different datasets and MTP problem settings identifies cases where specific HPO methods outperform others.
KW - Automated machine learning
KW - Hyperparameter optimization
KW - Machine learning
KW - Multi-label classification
KW - Multi-target prediction
UR - http://www.scopus.com/inward/record.url?scp=85200158240&partnerID=8YFLogxK
U2 - 10.1016/j.asoc.2024.111957
DO - 10.1016/j.asoc.2024.111957
M3 - Article
VL - 165
SP - 111957
JO - Applied soft computing
JF - Applied soft computing
SN - 1568-4946
M1 - 111957
ER -