ABSTRACT
Sensor networks are increasingly being used to provide timely information about the physical, urban and human environment. Algorithms that depend on sensor data often assume that the readings are complete. However, node failures or communication breakdowns result in missing data entries, preventing the use of such algorithms. To impute these missing values, we propose a method of exploiting spatial correlations which is based on the sparse autoencoder and inspired by the conditional Restricted Boltzmann Machine that contested for the Netflix Prize. We modify the autoencoder to cope with missing data, and test it on data from a sensor testbed in Santander, Spain. We show that our algorithm extracts features from datasets with high proportions of missing data and uses these features to accurately and efficiently impute missing entries.
- G. E. Hinton, N. Srivastava, A. Krizhevsky, I. Sutskever, and R. R. Salakhutdinov. Improving neural networks by preventing co-adaptation of feature detectors. arXiv preprint arXiv:1207.0580, 2012.Google Scholar
- L. Kong, M. Xia, X.-Y. Liu, M.-Y. Wu, and X. Liu. Data loss and reconstruction in sensor networks. In INFOCOM, 2013 Proceedings IEEE, pages 1654--1662. IEEE, 2013.Google ScholarCross Ref
- A. Ng. Sparse autoencoder. CS294A Lecture notes, page 72, 2011.Google Scholar
- L. Pan and J. Li. K-nearest neighbor based missing data estimation algorithm in wireless sensor networks. Wireless Sensor Network, 2(2), 2010.Google Scholar
- D. E. Rumelhart, G. E. Hinton, and R. J. Williams. Learning internal representations by error propagation. Technical report, DTIC Document, 1985.Google Scholar
- R. Salakhutdinov, A. Mnih, and G. Hinton. Restricted Boltzmann machines for collaborative filtering. In Proceedings of the 24th international conference on Machine learning, pages 791--798. ACM, 2007. Google ScholarDigital Library
- L. Sanchez, J. A. Galache, V. Gutierrez, J. Hernandez, J. Bernat, A. Gluhak, and T. Garcia. SmartSantander: The meeting point between future internet research and experimentation and the smart cities. In Future Network & Mobile Summit (FutureNetw), 2011, pages 1--8. IEEE, 2011.Google Scholar
- P. K. Sharpe and R. Solly. Dealing with missing values in neural network-based diagnostic systems. Neural Computing & Applications, 3(2):73--77, 1995.Google ScholarCross Ref
- P. Vamplew and A. Adams. Missing values in a backpropagation neural net. In Proceedings of the 3rd. Australian Conference on Neural Networks (ACNN), I, pages 64--66, 1992.Google Scholar
- L. Wan, M. Zeiler, S. Zhang, Y. L. Cun, and R. Fergus. Regularization of neural networks using DropConnect. In Proceedings of the 30th International Conference on Machine Learning (ICML-13), pages 1058--1066, 2013.Google ScholarDigital Library
Index Terms
- Imputing missing values in sensor networks using sparse data representations
Recommendations
Imputation of missing sensor data values using in-exact replicas
Sensor network systems attempt to exploit redundancy which is assumed to be inherent in a pervasive sensor network due to the density of the sensor nodes. However, there is a lack of sensor data stream correlation between seemingly redundant sensors. ...
A reinforcement learning-based approach for imputing missing data
AbstractMissing data is a major problem in real-world datasets, which hinders the performance of data analytics. Conventional data imputation schemes such as univariate single imputation replace missing values in each column with the same approximated ...
Using data mining to handle missing data in multi-hop sensor network applications
MobiDE '10: Proceedings of the Ninth ACM International Workshop on Data Engineering for Wireless and Mobile AccessA sensor's data loss or corruption, aka sensor data missing, is a common phenomenon in modern wireless sensor networks. It is more severe for multi-hop sensor network (MSN) applications where sensor data reach the base station via other sensors; hence a ...
Comments