Boosting model performance and interpretation by entangling preprocessing selection and variable selection
- Submitting institution
-
University of South Wales / Prifysgol De Cymru
- Unit of assessment
- 12 - Engineering
- Output identifier
- 1662591
- Type
- D - Journal article
- DOI
-
10.1016/j.aca.2016.08.022
- Title of journal
- Analytica Chimica Acta
- Article number
- -
- First page
- 44
- Volume
- 938
- Issue
- -
- ISSN
- 0003-2670
- Open access status
- Deposit exception
- Month of publication
- September
- Year of publication
- 2016
- 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
-
7
- Research group(s)
-
A - Sustainable Environment Research Centre
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- A Design of Experiments approach was applied for the first time to identify the best pre-processing strategies to make data model performance more robust. This international collaboration between industrialists and academics looked at the many data pre-processing techniques in use across a multitude of analytical techniques. Current best practice has been shown to be historical in nature or aimed to deliver better plots for manual interpretation but actual interfere with obtaining the best results when advanced multivariate chemometric techniques are applied. This paper lays the foundation for more robust data pre-processing strategies across a multitude of analytical techniques.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -