Heuristic target class selection for advancing performance of coverage-based rule learning
- Submitting institution
-
University of Portsmouth
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 14316699
- Type
- D - Journal article
- DOI
-
10.1016/j.ins.2018.12.001
- Title of journal
- Information Sciences
- Article number
- -
- First page
- 164
- Volume
- 479
- Issue
- -
- ISSN
- 0020-0255
- Open access status
- Compliant
- Month of publication
- December
- 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
-
2
- Research group(s)
-
B - Computational Intelligence
- Citation count
- 5
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- The proposed rule-based algorithm addresses the problem of complexity in previous algorithms while also improving the classification performance in comparison with a decision tree algorithm (C4.5), which is one of the most popular algorithms in machine learning. This research makes available a rule-based algorithm that leads to low-complexity, transparent and interpretable models, which are key for AI transparency and accountability – issues at the forefront of UK and EU agendas. The work was used for energy management in hybrid vehicles (Liu et al., Energy, 2020, 118212; Liu at al., J. Clean Production, 2020, 121017).
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -