Multi-Matrices Factorization with Application to Missing Sensor Data Imputation
2013 (English)In: Sensors, ISSN 1424-8220, Vol. 13, no 11, 15172-15186 p.Article in journal (Refereed) Published
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, T-1, T-2, . . . , T-t, where the entry, R-i,R-j, is the aggregate value of the data collected in the ith area at T-j. 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 epsilon R-dxt, where R-j approximate to U-(j)(T) T-j. 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.
Place, publisher, year, edition, pages
2013. Vol. 13, no 11, 15172-15186 p.
matrix factorization, sensor data, probabilistic graphical model, missing estimation
IdentifiersURN: urn:nbn:se:umu:diva-86631DOI: 10.3390/s131115172ISI: 000330321100049OAI: oai:DiVA.org:umu-86631DiVA: diva2:713843