Structure selection for convolutive non-negative matrix factorization using normalized maximum likelihood coding
- Submitting institution
-
University of Greenwich
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 30536
- Type
- E - Conference contribution
- DOI
-
10.1109/ICDM.2016.0163
- Title of conference / published proceedings
- 2016 IEEE 16th International Conference on Data Mining
- First page
- 1221
- Volume
- 0
- Issue
- -
- ISSN
- 2374-8486
- Open access status
- Deposit exception
- Month of publication
- -
- Year of publication
- 2017
- 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
-
2
- Research group(s)
-
-
- Citation count
- 1
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- This paper proposes a novel information criterion for convolutive nonnegative matrix factorisation (CNMF), which can be used to extract features on time-series data such as economic data and acoustic data. This information criterion is the first theoretical criterion for the CNMF. Owing to this criterion, we can select the rank of the factorisation, which is a vital hyperparameter of the CNMF. ICDM is IEEE International Conference on Data Mining. In 2016, the acceptance rate was 19.6%.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -
