Publication Target:
Annual Conference of the Prognostics and Health Management Society 2016
Submission Type:
Full Paper Attachment | Size | Timestamp |
---|---|---|
phmc_16_037.pdf | 997.4 KB | September 7, 2016 - 6:08pm |
This paper evaluates data-driven asset prognostic models from a modelling “ecosystem” perspective, which includes data description, uncertainty quantification, model selection justification and validation, and application limitations. An easily accessible and comprehensive ecosystem enables efficient reproducibility of previous work to facilitate both the adoption of the models by industry and the development of future scientific methods. The results of this study enable the development of a list of ecosystem elements to accompany the publication of new models. By describing the ecosystem in the communication of new models, researchers can ensure the reproducibility of their models in the wider prognostic community.
Publication Year:
2016
Publication Volume:
7
Publication Control Number:
037
Page Count:
9 Submission Keywords:
prognostics
Reproducibility
modelling ecosystem
Submission Topic Areas:
Standards and methodologies