In EDS ansehen

CL-NOTEARS: Continuous Optimization Algorithm Based on Curriculum Learning Framework

Gespeichert in:
Bibliographische Detailangaben
Titel: CL-NOTEARS: Continuous Optimization Algorithm Based on Curriculum Learning Framework
Autoren: Kaiyue Liu, Lihua Liu, Kaiming Xiao, Xuan Li, Hang Zhang, Yun Zhou, Hongbin Huang
Quelle: Mathematics, Vol 12, Iss 17, p 2640 (2024)
Verlagsinformationen: MDPI AG, 2024.
Publikationsjahr: 2024
Bestand: LCC:Mathematics
Schlagwörter: continuous optimization, Gaussian cluster, curriculum learning, casual structure, Mathematics, QA1-939
Beschreibung: Causal structure learning plays a crucial role in the current field of artificial intelligence, yet existing causal structure learning methods are susceptible to interference from data sample noise and often become trapped in local optima. To address these challenges, this paper introduces a continuous optimization algorithm based on the curriculum learning framework: CL-NOTEARS. The model utilizes the curriculum loss function during training as a priority evaluation metric for curriculum selection and formulates the sample learning sequence of the model through task-level curricula, thereby enhancing the model’s learning performance. A curriculum-based sample prioritization strategy is employed that dynamically adjusts the training sequence based on variations in loss function values across different samples throughout the training process. The results demonstrate a significant reduction in the impact of sample noise in the data, leading to improved model training performance.
Publikationsart: article
Dateibeschreibung: electronic resource
Sprache: English
ISSN: 2227-7390
Relation: https://www.mdpi.com/2227-7390/12/17/2640; https://doaj.org/toc/2227-7390
DOI: 10.3390/math12172640
Zugangs-URL: https://doaj.org/article/0d58c24f8c144008a4ab3f17f3fffd2e
Dokumentencode: edsdoj.0d58c24f8c144008a4ab3f17f3fffd2e
Datenbank: Directory of Open Access Journals
Beschreibung
Abstract:Causal structure learning plays a crucial role in the current field of artificial intelligence, yet existing causal structure learning methods are susceptible to interference from data sample noise and often become trapped in local optima. To address these challenges, this paper introduces a continuous optimization algorithm based on the curriculum learning framework: CL-NOTEARS. The model utilizes the curriculum loss function during training as a priority evaluation metric for curriculum selection and formulates the sample learning sequence of the model through task-level curricula, thereby enhancing the model’s learning performance. A curriculum-based sample prioritization strategy is employed that dynamically adjusts the training sequence based on variations in loss function values across different samples throughout the training process. The results demonstrate a significant reduction in the impact of sample noise in the data, leading to improved model training performance.
ISSN:22277390
DOI:10.3390/math12172640