Multi-Task Learning for Compositional Data via Sparse Network Lasso
Multi-task learning is a statistical methodology that aims to improve the generalization performances of estimation and prediction tasks by sharing common information among multiple tasks. On the other hand, compositional data consist of proportions as components summing to one. Because components o...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-12-01
|
| Series: | Entropy |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1099-4300/24/12/1839 |
| _version_ | 1797459184155885568 |
|---|---|
| author | Akira Okazaki Shuichi Kawano |
| author_facet | Akira Okazaki Shuichi Kawano |
| author_sort | Akira Okazaki |
| collection | DOAJ |
| description | Multi-task learning is a statistical methodology that aims to improve the generalization performances of estimation and prediction tasks by sharing common information among multiple tasks. On the other hand, compositional data consist of proportions as components summing to one. Because components of compositional data depend on each other, existing methods for multi-task learning cannot be directly applied to them. In the framework of multi-task learning, a network lasso regularization enables us to consider each sample as a single task and construct different models for each one. In this paper, we propose a multi-task learning method for compositional data using a sparse network lasso. We focus on a symmetric form of the log-contrast model, which is a regression model with compositional covariates. Our proposed method enables us to extract latent clusters and relevant variables for compositional data by considering relationships among samples. The effectiveness of the proposed method is evaluated through simulation studies and application to gut microbiome data. Both results show that the prediction accuracy of our proposed method is better than existing methods when information about relationships among samples is appropriately obtained. |
| first_indexed | 2024-03-09T16:47:46Z |
| format | Article |
| id | doaj.art-4ce7756a6a054d6b9c0ac77dc27f5baf |
| institution | Directory Open Access Journal |
| issn | 1099-4300 |
| language | English |
| last_indexed | 2024-03-09T16:47:46Z |
| publishDate | 2022-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Entropy |
| spelling | doaj.art-4ce7756a6a054d6b9c0ac77dc27f5baf2023-11-24T14:43:56ZengMDPI AGEntropy1099-43002022-12-012412183910.3390/e24121839Multi-Task Learning for Compositional Data via Sparse Network LassoAkira Okazaki0Shuichi Kawano1Graduate School of Informatics and Engineering, The University of Electro-Communications, 1-5-1 Chofugaoka, Chofu 182-8585, Tokyo, JapanFaculty of Mathematics, Kyushu University, 744 Motooka, Nishi-ku 819-0395, Fukuoka, JapanMulti-task learning is a statistical methodology that aims to improve the generalization performances of estimation and prediction tasks by sharing common information among multiple tasks. On the other hand, compositional data consist of proportions as components summing to one. Because components of compositional data depend on each other, existing methods for multi-task learning cannot be directly applied to them. In the framework of multi-task learning, a network lasso regularization enables us to consider each sample as a single task and construct different models for each one. In this paper, we propose a multi-task learning method for compositional data using a sparse network lasso. We focus on a symmetric form of the log-contrast model, which is a regression model with compositional covariates. Our proposed method enables us to extract latent clusters and relevant variables for compositional data by considering relationships among samples. The effectiveness of the proposed method is evaluated through simulation studies and application to gut microbiome data. Both results show that the prediction accuracy of our proposed method is better than existing methods when information about relationships among samples is appropriately obtained.https://www.mdpi.com/1099-4300/24/12/1839clusteringlog-contrast modelmulti-task learningsymmetric formvariable selection |
| spellingShingle | Akira Okazaki Shuichi Kawano Multi-Task Learning for Compositional Data via Sparse Network Lasso Entropy clustering log-contrast model multi-task learning symmetric form variable selection |
| title | Multi-Task Learning for Compositional Data via Sparse Network Lasso |
| title_full | Multi-Task Learning for Compositional Data via Sparse Network Lasso |
| title_fullStr | Multi-Task Learning for Compositional Data via Sparse Network Lasso |
| title_full_unstemmed | Multi-Task Learning for Compositional Data via Sparse Network Lasso |
| title_short | Multi-Task Learning for Compositional Data via Sparse Network Lasso |
| title_sort | multi task learning for compositional data via sparse network lasso |
| topic | clustering log-contrast model multi-task learning symmetric form variable selection |
| url | https://www.mdpi.com/1099-4300/24/12/1839 |
| work_keys_str_mv | AT akiraokazaki multitasklearningforcompositionaldataviasparsenetworklasso AT shuichikawano multitasklearningforcompositionaldataviasparsenetworklasso |