Feature Extraction for Incomplete Data Via Low-Rank Tensor Decomposition With Feature Regularization
- Submitting institution
-
The University of Sheffield
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 2512
- Type
- D - Journal article
- DOI
-
10.1109/TNNLS.2018.2873655
- Title of journal
- IEEE Transactions on Neural Networks and Learning Systems
- Article number
- -
- First page
- 1803
- Volume
- 30
- Issue
- 6
- ISSN
- 2162-237X
- Open access status
- Compliant
- Month of publication
- October
- Year of publication
- 2018
- URL
-
-
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
3
- Research group(s)
-
C - Machine Learning
- Citation count
- 8
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- This is the first paper to tackle the problem of direct feature extraction from incomplete tensors with two novel methods proposed under a unifying unsupervised learning framework. The work is significant because it provides a way to enable extraction of informative features directly from incomplete multidimensional data, a common real-world challenging problem limiting the utilisation of precious data sources. It led to a grant of £639,873 awarded by the Wellcome Trust (215799/Z/19/Z, collaboration with NHS, contact: Director of the Sheffield Pulmonary Vascular Disease Unit) as a key technology to deal with missing entries in electronic health records.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -