|Reference : Design Space and desirability index. A Bayesian predictive risk-based approach to fle...|
|Scientific congresses and symposiums : Unpublished conference/Abstract|
|Physical, chemical, mathematical & earth Sciences : Mathematics|
|Design Space and desirability index. A Bayesian predictive risk-based approach to flexibly achieve multi-criteria decision methods.|
|Lebrun, Pierre [Université de Liège - ULg > Département de pharmacie > Chimie analytique >]|
|Boulanger, Bruno [ > > ]|
|Hubert, Philippe [Université de Liège - ULg > Département de pharmacie > Chimie analytique >]|
|Mbinze Kindenge, Jérémie [Université de Liège - ULg > > > Form. doc. sc. bioméd. & pharma.]|
|Debrus, Benjamin [Université de Liège - ULg > Département de pharmacie > Chimie analytique >]|
|The Second International Symposium on Biopharmaceutical Statistics|
|from 1-3-2011 to 3-3-2011|
|The International Society for Biopharmaceutical Statistics|
|[en] The Design Space (DS) is defined as the set of factors settings (input conditions) that will provide results at least better than pre-defined acceptance limits. The proposed methodology aims at identifying a region in the space of factors that will likely provide satisfactory results during the future use of an analytical method or process in routine, through an optimization process.
In a Bayesian framework, the responses are modelled using a multivariate multiple regression model allowing deriving their joint predictive posterior distribution.
On the basis of this consequent distribution, a multi-criteria risk-based decision is taken with respect to the pre-defined acceptance limits. This aims to identify the DS. In this context, desirability methodologies are also applied to take the risk-based decision in a more flexible way.
An example based on high-performance liquid chromatography illustrates the applicability of the methodology with highly correlated and constrained responses.
|Researchers ; Professionals ; Students|
|File(s) associated to this reference|
All documents in ORBi are protected by a user license.