PERIOD: MAY - JUNE - JULY 2021 The course presents Chemometrics as a set of tools for experimental data acquisition, preprocessing, exploration and analysis aimed at optimizing and maximizing the extraction of valuable physico-chemical information through a multivariate approach. The course consists of both frontal lessons, during which the chemometric techniques will be introduced, and practical sessions in which the students will be challenged on case studies proposed either by the lecturers or by the students themselves.

Matlab, Excel, math. analysis, fundaments of linear algebra

The course will consist of a theoretical part with frontal lessons and a practical part strongly focused on problem solving, during which the students will be challenged on a dataset from which information has to be extracted using the chemometric techniques explained in the course. Introduction: What is and what is for Chemometrics?! Design of experiment (DoE): why are experiments performed, how to plan the experiments to optimize the amount of information that can be extracted, the experimental domain, why changing “one variable at the time” not always is the best approach, experimental designs and response surfaces, examples of DoE techniques. Data inspection and preprocessing: visual inspection, data pre-processing and pre-treatment. Exploratory analysis: Principal Component Analysis (PCA), clustering analysis, residuals inspection, outlier detection, rank of a dataset, overfit and underfit, the importance of metadata. Regression and classification: multivariate regression methods (PLS, Partial Least Squares), multivariate classification methods (PLS-DA, Partial Least Squares-Discriminant Analysis). Case study: application of the proposed chemometric techniques to a real-life dataset proposed either by the lecturers or the students themselves.

On site

Oral presentation - Team project work development

P.D.2-2 - June