Hyperparameter selection for dataset‐constrained semantic segmentation: Practical machine learning optimization
The proposed optimization method gives a systematic and quantitative approach to something intuitively understood, that hyperparameters change model performance. Even just grid search optimization with random forest analysis presented here can be informative within hardware and data quality/availability limitations, adding confidence to model validity and minimize decision-making risks. By providing a guided methodology, this work helps medical physicists to improve their model optimization, irrespective of specific challenges posed by datasets and model design.
Authors
- Mark Jenkinson (ORCID: https://orcid.org/0000-0001-6043-0166)
- Timothy John Kleinig (ORCID: https://orcid.org/0000-0003-4430-3276)
- Eva Bezak (ORCID: https://orcid.org/0000-0002-1315-1735)
- Wolfgang Mayer (ORCID: https://orcid.org/0000-0002-2154-2269)
- Joseph Dawson (ORCID: https://orcid.org/0000-0001-9229-9545)
- G. Brown (ORCID: https://orcid.org/0000-0001-7994-2743)
- Chris Boyd (ORCID: https://orcid.org/0000-0001-6443-0147)
Institutions
- Royal Adelaide Hospital (AU)
- University of South Australia (AU)
- University of Oxford (GB)
- Wellcome Centre for Integrative Neuroimaging (GB)
- Australian Centre for Robotic Vision (AU)
- South Australian Health and Medical Research Institute (AU)
- The University of Adelaide (AU)
- Wellcome Trust (GB)
Publication Details
- Journal
- Journal of Applied Clinical Medical Physics
- Published
- 2024-10-10
- DOI
- https://doi.org/10.1002/acm2.14542
- Citations
- 6
- Primary Topic
- Machine Learning and Data Classification
- Type
- article
- Field-Weighted Citation Impact
- 2.08