NSCPHD1191 Copenhagen School of Chemometrics - lphd191
1. DOE. Introduction to Design of Experiments
The basic theory and practice of Design of Experiments and statistical inference is revisited. The aim is to give a practitioner idea or reminder about the main features and uses of DoE. The seminar will be based on teaching hours and some guided exercises.
2. PCA. Principal Component Analysis as the most versatile tool in Data Analysis
Principal Component Analysis has become in the most powerful and versatile tool for analyzing data tables in Analytical Sciences. Here we present a course to show the main benefits and drawbacks of PCA when it is used for different kind of analytical data: Spectroscopy, environmental assessment, experiments performance, etc.
3. CLASS I. Supervised classification methods (I). Linear models. SIMCA and PLS-DA
The seminar is focused on the theory and practice of two linear classification tools (SIMCA and PLS-DA). The seminar will be based on teaching hours with guided exercises and practical sessions with real cases.
4. CLASS II. Supervised classification methods (II). Non-linear models. ANN and SVM
The seminar is focused on the theory and practice of advanced non-linear classification methods for Pattern Recognition, to give a practitioner idea about their main features and uses. The seminar will be based on teaching hours with guided exercises and practical sessions with real cases.
5. MCR. Multivariate Curve Resolution. Advanced deconvolution of signals.
In the last years, curve resolution techniques are gaining importance in the modeling of different analytical data. Especially, Multivariate Curve Resolution has widely demonstrated its usefulness in kinetic modeling, solving problems in chromatographic data (peak resolution/deconvolution) and hyperspectral images.
6. CHROM I. Chromatographic Data Analysis (I). Baseline correction, peak alignment and normalization
The seminar is focused on solving three major issues in chromatographic data: Retention time shifts between different samples, handling of baseline drifts and peak normalization techniques.
7. CHROM II. Chromatographic Data Analysis (II). Fingerprinting
The seminar is focused on interpreting chromatographic data and the results derived from fingerprinting analysis (pattern recognition/Principal Component Analysis).
8. CHROM III. Chromatographic Data Analysis (III). Curve resolution/advanced deconvolution of overlapped peaks
The seminar is focused on solving the problem of overlapped peaks when a hyphenated chromatographic device is used (e.g. HPLC-DAD, GC-MS, etc. Moreover, several more issues of chromatographic data (e.g. baseline drifts or retention time shifts between samples) will be handled by using curve resolution/advanced peak deconvolution methods like MCR and PARAFAC2.
9. SPECT I. Spectroscopic data analysis (I). Unsupervised Pattern recognition (PCA)
The first seminar of the series focused on spectroscopic data analysis will be based on 1) pre-processing techniques of spectral data (mainly NIR, Raman and UV-Vis) and 2) develop Principal Component Analysis models in spectral data.
10. SPECT II. Spectroscopic data analysis (II). Regression models (PLS)
The second seminar of the series focused on spectroscopic data analysis will be based on the development of calibration models based on Partial Least Squares Regression in spectral data (mainly NIR, Raman and UV-Vis). This includes: Calibration development, validation, variable selection, etc.
11. STAT. Validation of models. Statistical inference in Multivariate Data Analysis.
In this seminar, extremely relevant topics like validation of multivariate models, figures of merits and statistical parameters to assess the robustness and precision of multivariate models will be taught. This seminar is highly recommended for those who want to learn how to assess a multivariate model in a statistical manner and those ones attending seminars CLASS I, CLASS II, MCR, CHROM I, II, III, SPECT I, II.
One of the key points of CSC is the interaction between the students and between the teachers to offer the possibility of open-mind discussion forums always within the framework of scientific data analysis and performance. That is why CSC counts with well-recognized experts on chemometrics and multivariate data analysis in their respective fields. The intended learning outcomes have to be divided into two categories:
- Individual learning outcomes: the main target for each individual seminar is to learn the basis of one data analysis method focused on several proposed examples. The student has to be able to apply the acquired knowledge to any problem related to the seminar by himself.
- Global learning outcomes: The students attending all the seminars, at the end of the course they will be able to understand the structure of a huge amount of data structures and also to understand the problems derived from the data. Moreover, they will be independent in the application of solutions to their problems in a dedicated manner.
- Practical exercises
- Project work
- Theory exercises
To sign up for the course, please send
an e-mail to Jeanette V. Hansen (firstname.lastname@example.org) or to José Manuel
Amigo (email@example.com) with the following information: Name:
Position/department: University: Seminars that you want to attend:
The deadline for registration to any of the seminars will be the 1st of September of 2013.
- 12 ECTS
- Type of assessment
- Written assignmentWritten reports; work out by the participants individually. The reports have to be handed in maximally one week after the last lecture and are evaluated and credited with PASS/FAIL by the course lectures.
- Exam registration requirements
- 1 ECTS per seminar. The time for reports is dedicated to consultancy and supervision by the student having an importance of 1 ECTS. One student attending all the seminars, he will get 12 ECTS.