Seminário "Tackling problems in spectroscopy and chemometrics with AI"

-
Campus de Gambelas - Sala 2.11 da Faculdade de Ciências e Tecnologia, edifício 2
Share

No dia 6 de dezembro, pelas 11h30, irá realizar-se o seminário "Tackling problems in spectroscopy and chemometrics with Artificial Intelligence", que terá lugar na sala 2.11 da Faculdade de Ciências e Tecnologia, edifício 2 do Campus de Gambelas.

Inserido no Ciclo de Seminários do Departamento de Física intitulado "Conversas de Física", o seminário será apresentado por Dário Passos, investigador do Centro de Eletrónica, Optoeletrónica e Telecomunicações (CEOT) da UAlg.

A entrada é livre.

Resumo

"Vis-NIR spectroscopy is nowadays one of the most used techniques to study the physicochemical properties of biological tissues. _e idea is to extract information from the non-destructive interaction between light and tissue to gain knowledge about its properties. Practical examples are oxygenation levels in human blood or internal quality parameters of fruit. Machine Learning (ML) algorithms are a fundamental tool in the computational pipeline necessary to map the spectral information to the properties of interest. Over the last decade, a particular type of ML technique known as Deep Learning has revolutionized computer vision and natural language processing tasks. In this talk, I'll briefly introduce the concept of Deep Learning, how it can be applied to multiple spectroscopy and chemometrics problems and how it has the potential to enhance computational physics."