Improving Neural Network-Based Multi-Label Classification With Pattern Loss Penalties
This research work introduces two novel loss functions, pattern-loss (POL) and label similarity-based instance modeling (LSIM), for improving the performance of multi-label classification using artificial neural network-based techniques. These loss functions incorporate additional optimization const...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2024-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10495049/ |
| _version_ | 1797206105390055424 |
|---|---|
| author | Worawith Sangkatip Phatthanaphong Chomphuwiset Kaveepoj Bunluewong Sakorn Mekruksavanich Emmanuel Okafor Olarik Surinta |
| author_facet | Worawith Sangkatip Phatthanaphong Chomphuwiset Kaveepoj Bunluewong Sakorn Mekruksavanich Emmanuel Okafor Olarik Surinta |
| author_sort | Worawith Sangkatip |
| collection | DOAJ |
| description | This research work introduces two novel loss functions, pattern-loss (POL) and label similarity-based instance modeling (LSIM), for improving the performance of multi-label classification using artificial neural network-based techniques. These loss functions incorporate additional optimization constraints based on the distribution of multi-label class patterns and the similarity of data instances. By integrating these patterns during the network training process, the trained model is tuned to align with the existing patterns in the training data. The proposed approach decomposes the loss function into two components: the cross entropy loss and the pattern loss derived from the distribution of class-label patterns. Experimental evaluations were conducted on eight standard datasets, comparing the proposed methods with three existing techniques.The results demonstrate the effectiveness of the proposed approach, with POL and LSIM consistently achieving superior accuracy performance compared to the benchmark methods. |
| first_indexed | 2024-04-24T09:01:43Z |
| format | Article |
| id | doaj.art-4eecc0f0f8034f7d8a980a89f299dbe7 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-04-24T09:01:43Z |
| publishDate | 2024-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-4eecc0f0f8034f7d8a980a89f299dbe72024-04-15T23:00:53ZengIEEEIEEE Access2169-35362024-01-0112522375224810.1109/ACCESS.2024.338684110495049Improving Neural Network-Based Multi-Label Classification With Pattern Loss PenaltiesWorawith Sangkatip0Phatthanaphong Chomphuwiset1Kaveepoj Bunluewong2https://orcid.org/0009-0007-8925-4160Sakorn Mekruksavanich3https://orcid.org/0000-0002-3735-4262Emmanuel Okafor4Olarik Surinta5https://orcid.org/0000-0002-0644-1435Department of Information Technology, Faculty of Informatics, Mahasarakham University, Kham Riang, ThailandDepartment of Computer Science, Faculty of Informatics, Polar Laboratory, Mahasarakham University, Kham Riang, ThailandDepartment of Computer Science, Faculty of Informatics, Polar Laboratory, Mahasarakham University, Kham Riang, ThailandDepartment of Computer Engineering, School of Information and Communication Technology, University of Phayao, Mae Ka, ThailandSDAIA-KFUPM Joint Research Center for Artificial Intelligence, King Fahd University of Petroleum and Minerals, Dhahran, Saudi ArabiaDepartment of Information Technology, Faculty of Informatics, Multi-Agent Intelligent Simulation Laboratory (MISL) Research Unit, Mahasarakham University, Kham Riang, ThailandThis research work introduces two novel loss functions, pattern-loss (POL) and label similarity-based instance modeling (LSIM), for improving the performance of multi-label classification using artificial neural network-based techniques. These loss functions incorporate additional optimization constraints based on the distribution of multi-label class patterns and the similarity of data instances. By integrating these patterns during the network training process, the trained model is tuned to align with the existing patterns in the training data. The proposed approach decomposes the loss function into two components: the cross entropy loss and the pattern loss derived from the distribution of class-label patterns. Experimental evaluations were conducted on eight standard datasets, comparing the proposed methods with three existing techniques.The results demonstrate the effectiveness of the proposed approach, with POL and LSIM consistently achieving superior accuracy performance compared to the benchmark methods.https://ieeexplore.ieee.org/document/10495049/Multi-label classificationlabel correlationlabel-specific featuresdeep neural networkloss functions |
| spellingShingle | Worawith Sangkatip Phatthanaphong Chomphuwiset Kaveepoj Bunluewong Sakorn Mekruksavanich Emmanuel Okafor Olarik Surinta Improving Neural Network-Based Multi-Label Classification With Pattern Loss Penalties IEEE Access Multi-label classification label correlation label-specific features deep neural network loss functions |
| title | Improving Neural Network-Based Multi-Label Classification With Pattern Loss Penalties |
| title_full | Improving Neural Network-Based Multi-Label Classification With Pattern Loss Penalties |
| title_fullStr | Improving Neural Network-Based Multi-Label Classification With Pattern Loss Penalties |
| title_full_unstemmed | Improving Neural Network-Based Multi-Label Classification With Pattern Loss Penalties |
| title_short | Improving Neural Network-Based Multi-Label Classification With Pattern Loss Penalties |
| title_sort | improving neural network based multi label classification with pattern loss penalties |
| topic | Multi-label classification label correlation label-specific features deep neural network loss functions |
| url | https://ieeexplore.ieee.org/document/10495049/ |
| work_keys_str_mv | AT worawithsangkatip improvingneuralnetworkbasedmultilabelclassificationwithpatternlosspenalties AT phatthanaphongchomphuwiset improvingneuralnetworkbasedmultilabelclassificationwithpatternlosspenalties AT kaveepojbunluewong improvingneuralnetworkbasedmultilabelclassificationwithpatternlosspenalties AT sakornmekruksavanich improvingneuralnetworkbasedmultilabelclassificationwithpatternlosspenalties AT emmanuelokafor improvingneuralnetworkbasedmultilabelclassificationwithpatternlosspenalties AT olariksurinta improvingneuralnetworkbasedmultilabelclassificationwithpatternlosspenalties |