Chemometrics

Kemometria

4 credits – 2 contact hours weekly

code: chemometk20em (En/Hu)

programme: MSc in Chemistry from year 2020/21, status: semi-optional

Lecturer: Gergely Tóth

The course is integrated to Moodle

Language: English/Hungarian*

*(spring semester 2020/21: officially it is Hungarian, but on request I switch to English. Online materials, books and the Moodle-study module are in English or in both languages)

Aim of the course: The aim of the course is to overview the basic methods used in uni- and multivariate data analysis on chemical data, including mathematical background. The project parts contain examples on the use of the methods on analytical and pharmaceutical data. The course is recommended for future chemist evaluating data, especially in analytics, and for jobs were data communication with statistics is required.**

**(This new MSc in Chemistry course integrates my previous chemometric course and more project elements)

Lecture part

Online lectures weekly, if the number of active students is at least 5. or

Online consultations-tutorials every second week, if there are less students.

Topics: basics of univariate probability theory and statistics, basic distributions, some tests, multivariate data and linear spaces, distances, clustering, classification, principal component analysis, factor analysis, linear regression, biased and advanced linear regression, model validation, non-linear modelling, multivariate curve resolution, methods of variance analysis, data transformations

Project part

Two datasets with the corresponding articles are obtained, where data analysis have to be performed, e.g., by Matlab, and two reports have to be written (6-12 pages long ones)

Assessment

25% data evaluation report I

25% data evaluation report II

25% self-study tests in Moodle

25% final test in Moodle

The results in parts A, B, C and D will be averaged. The final mark is calculated as (1)<40%<=(2)<55%<=(3)<70%<=(4)<85%<=(5). For grading it is required to show a minimal activity on all parts.

Workload in hours

24 - lecture

32 – self-study on theory combined with the self-study tests

2 – consultation for lecture

2 – exam

2*26 – data evaluation using Matlab, report writing

8 – consultation for reports

sum: 120 hours