Deep learning in data analysis for the classification and selection of biomass valorization routes

The industry has a high contrast between waste production and profits. Because of it, the use of biomass is an important process. However, its process is long overdue considering the time to extract the biomass, the characterization, and finally the long and privileged search to find ideas in the routes of use. The present work gives the opportunity to have consolidated the principal information about the uses of principal biomass and, after it, obtain a general view of each type of biomass for the characterization protocol used (NREL, PA, or UA). Finally, it has three excellent machine learning models that allow the categorization of each biomass into different groups, made from the unsupervised learning of K-means, deeply studied to make it as unbiased as possible. The models have more than 90% excellence in their process, so they are reliable to the industry and academic community. Its thesis shows the real relationship between the use of new technologies and the chemical process, which so far has very few and only requires an analytical view.

• Autores: Suárez Díaz, Yuli Natalia
• Tipo de recurso: Trabajo de grado de pregrado
• Fecha de publicación: 2022
• Institución: Universidad de los Andes
• Repositorio: Séneca: repositorio Uniandes
• Idioma: eng
• OAI Identifier: oai:repositorio.uniandes.edu.co:1992/58621
• Acceso en línea: http://hdl.handle.net/1992/58621
• Palabra clave: Deep learning; Biomass; Protocol; Biomasa; Ingeniería
• Rights: openAccess
• License: Attribution-NonCommercial-NoDerivatives 4.0 Internacional