Data clustering for classification of vegetable biomass from compositional data
A tool for biomass valorization
D. D. Durán-Aranguren (Technical University of Denmark (DTU))
Juan Toro-Delgado (Universidad de los Andes)
Valentina Núñez-Barrero (Universidad de los Andes)
Valentina Florez-Bulla (Universidad de los Andes)
R. Sierra (Universidad de los Andes)
J.A. Posada Duque (TU Delft - BT/Biotechnology and Society)
Solange I. Mussatto (Technical University of Denmark (DTU))
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
Compositional data on vegetable biomass is widely available from research papers and online databases. However, the high diversity of biomass characteristics and composition represents a challenge for researchers and companies willing to produce novel substances from residues, and that should decide on the best and most feasible options for their use as feedstocks. The present study constructed a database with information gathered from the proximate, ultimate, and chemical composition of different biomass residues that can be used for data analysis and classification to elucidate better how they can be valorized. Different data clustering techniques were implemented to determine how compositional data can be segmented. The identified groups, that contained residues with similar characteristics, allowed to have an insight into the valorization of these biomasses, which can be used as an initial tool for biorefinery design. The use of data clustering facilitated the identification of different types of biomasses in a systematic way, which until now has not been reported in the literature.
help_outline