Imperial College London
Portrait Picture

Dr Joram M. Posma PhD MSc B AS MRSC

Faculty of Medicine, Department of Metabolism, Digestion and Reproduction

Senior Lecturer in Biomedical Informatics
 
 
 
//

Contact

 

j.posma11 Website

 
 
//

Location

 

E305Burlington DanesHammersmith Campus

//

Summary

 

Publications

Citation

BibTex format

@inbook{Posma:2018:10.1016/B978-0-12-812293-8.00009-8,
author = {Posma, JM},
booktitle = {The Handbook of Metabolic Phenotyping},
doi = {10.1016/B978-0-12-812293-8.00009-8},
editor = {Lindon and Holmes and Nicholson},
publisher = {Elsevier},
title = {Multivariate statistical methods for metabolic phenotyping},
url = {http://dx.doi.org/10.1016/B978-0-12-812293-8.00009-8},
year = {2018}
}
Download

RIS format (EndNote, RefMan)

TY  - CHAP
AB  - The nature of metabolic phenotyping data, where typically more variables are measured than samples are available, requires careful application of statistical methods in order to be able to make meaningful inferences from the data. This Chapter describes different aspects of the multivariate modelling of this type of data, including data transformations and partitioning, unsupervised algorithms for dimension reduction, supervised algorithms for classification, clustering and regression, metrics and methods for obtaining unbiased prediction error estimates of predictive models and statistical spectroscopy tools used for biomarker identification. It focusses on describing methods routinely applied in the field as well as discussing methods that, as computational advancements are made, are poised to become more widely applied to metabolic phenotyping data.
AU  - Posma,JM
DO  - 10.1016/B978-0-12-812293-8.00009-8
PB  - Elsevier
PY  - 2018///
TI  - Multivariate statistical methods for metabolic phenotyping
T1  - The Handbook of Metabolic Phenotyping
UR  - http://dx.doi.org/10.1016/B978-0-12-812293-8.00009-8
ER  -
Download