Multi-Matrices Factorization with Application to Missing Sensor Data Imputation
We formulate a multi-matrices factorization model (MMF) for the missing sensor data estimation problem. The estimation problem is adequately transformed into a matrix completion one. With MMF, an n-by-t real matrix, R, is adopted to represent the data collected by mobile sensors from n areas at the...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2013-11-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | http://www.mdpi.com/1424-8220/13/11/15172 |
| _version_ | 1798006846509809664 |
|---|---|
| author | Wen-Xue Cai Rong Pan Lei Li Kang Chen Xian-Hong Xiang Wubin Li Xiao-Yu Huang |
| author_facet | Wen-Xue Cai Rong Pan Lei Li Kang Chen Xian-Hong Xiang Wubin Li Xiao-Yu Huang |
| author_sort | Wen-Xue Cai |
| collection | DOAJ |
| description | We formulate a multi-matrices factorization model (MMF) for the missing sensor data estimation problem. The estimation problem is adequately transformed into a matrix completion one. With MMF, an n-by-t real matrix, R, is adopted to represent the data collected by mobile sensors from n areas at the time, T1, T2, ... , Tt, where the entry, Rij, is the aggregate value of the data collected in the ith area at Tj . We propose to approximate R by seeking a family of d-by-n probabilistic spatial feature matrices, U(1), U(2), ... , U(t), and a probabilistic temporal feature matrix, V E Rdxt, where Rj ≈ UT(j)Tj . We also present a solution algorithm to the proposed model. We evaluate MMF with synthetic data and a real-world sensor dataset extensively. Experimental results demonstrate that our approach outperforms the state-of-the-art comparison algorithms. |
| first_indexed | 2024-04-11T13:01:07Z |
| format | Article |
| id | doaj.art-f3b281a90c7e4cbca0090a90eea413b2 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-04-11T13:01:07Z |
| publishDate | 2013-11-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-f3b281a90c7e4cbca0090a90eea413b22022-12-22T04:22:56ZengMDPI AGSensors1424-82202013-11-011311151721518610.3390/s131115172Multi-Matrices Factorization with Application to Missing Sensor Data ImputationWen-Xue CaiRong PanLei LiKang ChenXian-Hong XiangWubin LiXiao-Yu HuangWe formulate a multi-matrices factorization model (MMF) for the missing sensor data estimation problem. The estimation problem is adequately transformed into a matrix completion one. With MMF, an n-by-t real matrix, R, is adopted to represent the data collected by mobile sensors from n areas at the time, T1, T2, ... , Tt, where the entry, Rij, is the aggregate value of the data collected in the ith area at Tj . We propose to approximate R by seeking a family of d-by-n probabilistic spatial feature matrices, U(1), U(2), ... , U(t), and a probabilistic temporal feature matrix, V E Rdxt, where Rj ≈ UT(j)Tj . We also present a solution algorithm to the proposed model. We evaluate MMF with synthetic data and a real-world sensor dataset extensively. Experimental results demonstrate that our approach outperforms the state-of-the-art comparison algorithms.http://www.mdpi.com/1424-8220/13/11/15172matrix factorizationsensor dataprobabilistic graphical modelmissing estimation |
| spellingShingle | Wen-Xue Cai Rong Pan Lei Li Kang Chen Xian-Hong Xiang Wubin Li Xiao-Yu Huang Multi-Matrices Factorization with Application to Missing Sensor Data Imputation Sensors matrix factorization sensor data probabilistic graphical model missing estimation |
| title | Multi-Matrices Factorization with Application to Missing Sensor Data Imputation |
| title_full | Multi-Matrices Factorization with Application to Missing Sensor Data Imputation |
| title_fullStr | Multi-Matrices Factorization with Application to Missing Sensor Data Imputation |
| title_full_unstemmed | Multi-Matrices Factorization with Application to Missing Sensor Data Imputation |
| title_short | Multi-Matrices Factorization with Application to Missing Sensor Data Imputation |
| title_sort | multi matrices factorization with application to missing sensor data imputation |
| topic | matrix factorization sensor data probabilistic graphical model missing estimation |
| url | http://www.mdpi.com/1424-8220/13/11/15172 |
| work_keys_str_mv | AT wenxuecai multimatricesfactorizationwithapplicationtomissingsensordataimputation AT rongpan multimatricesfactorizationwithapplicationtomissingsensordataimputation AT leili multimatricesfactorizationwithapplicationtomissingsensordataimputation AT kangchen multimatricesfactorizationwithapplicationtomissingsensordataimputation AT xianhongxiang multimatricesfactorizationwithapplicationtomissingsensordataimputation AT wubinli multimatricesfactorizationwithapplicationtomissingsensordataimputation AT xiaoyuhuang multimatricesfactorizationwithapplicationtomissingsensordataimputation |