Interactive feature space extension for multidimensional data projection
- Submitting institution
-
Middlesex University
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 5
- Type
- D - Journal article
- DOI
-
10.1016/j.neucom.2014.09.061
- Title of journal
- Neurocomputing
- Article number
- -
- First page
- 611
- Volume
- 150
- Issue
- Part B
- ISSN
- 0925-2312
- Open access status
- Out of scope for open access requirements
- Month of publication
- October
- Year of publication
- 2014
- URL
-
http://eprints.mdx.ac.uk/20765/
- 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
-
5
- Research group(s)
-
-
- Citation count
- 5
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- This paper proposes a novel technique for projecting multi-dimensional data to a 2D/3D display as scatter-plots for pattern analysis. The work is significant because it improves traditional projection techniques by providing clearer group separation in the visualization so that patterns can be perceived more easily. The paper proposes different transformation strategies for improving group separation in the visualization. Experimental results show that a good trade-off can be achieved by applying different strategies to assure the visualization not only has good group separation, but also accurately preserves the original structure of the data.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -