Visual analytics for collaborative human-machine confidence in human-centric active learning tasks
- Submitting institution
-
University of the West of England, Bristol
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 854799
- Type
- D - Journal article
- DOI
-
10.1186/s13673-019-0167-8
- Title of journal
- Human-Centric Computing and Information Sciences
- Article number
- 5
- First page
- -
- Volume
- 9
- Issue
- -
- ISSN
- 2192-1962
- Open access status
- Compliant
- Month of publication
- February
- Year of publication
- 2019
- URL
-
https://doi.org/10.1186/s13673-019-0167-8
- 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
- 2
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- This paper describes a novel approach to explainable machine learning using visual analytics, to greater support human understanding of how the model is performing, whilst also exploring how a machine learns from a human based on their interactions and their confidence. Related to active learning, we demonstrate how our approach can dramatically improve classifier performance on common ML datasets compared to traditional batch learning methods, whilst also visually highlighting knowledge limitations in the machine model. The work was funded by the Defence Science and Technology Laboratory (DSTL) and has attracted attention from the Royal Air Force for future projects.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -
