This course focuses on the validation of Partial Least Squares (PLS) models, specifically on selecting the number of components. The course covers topics such as outliers, regression error measures, predicted versus measured plots, and cross-validation techniques. By the end of the course, learners will be able to validate PLS models effectively and make informed decisions on the number of components to use. The intended audience for this course includes individuals interested in chemometrics, machine learning, and data analysis.
Overview
Syllabus
Validation of PLS models - always important!
Outliers - why, how and when...?
Regression - Error measures
Predicted versus Measured plot
Cross validation - being smart with segments • Chemical analyses by six laboratories
Cross validation in more detail!
Secret trick - the other thing cross-validation does
Choice is not always simple - A few rules of thumb Rule
Taught by
Chemometrics & Machine Learning in Copenhagen
