Chapter 8 - Data Fusion by Multivariate Curve Resolution

Abstract

Abstract Data fusion implies often the concatenation of data sets that present an enormous diversity in terms of information, size, and behavior. The pieces of information connected reflect the variation apportioned by components, events, or sources that are differently represented and, yet, complement each other in the data blocks analyzed simultaneously. Multivariate curve resolution (MCR) was born as a tool to unmix the information in a single data set into a bilinear model of chemically meaningful profiles associated with pure components or sources. With the increase of complexity of chemical problems and the need to perform data fusion to understand all the aspects related to a particular scenario, multiset analysis by MCR came into play. Multiset analysis performed by MCR has two main advantages, the first stemming from the intrinsic versatile multiset structure and the second linked to the per block, per component, and per mode flexible application of constraints to model pure profiles by MCR, which covers the specific needs of the diverse blocks of information present in a data fusion framework. These two essential aspects are extensively developed in this chapter, and a final representative report on the main fields of application of data fusion by MCR is also provided.