|
Lecturer(s)
|
|
|
|
Course content
|
Topics of lectures after weeks of semester 1. Experiment planning, principle and reasons for use, different methods. 2. Multidimensional data, data matrix, objects and variables. Variable types and multidimensional random vector. 3. Multidimensional data preprocessing: Transformation types. Data centering and standardization. 4. Exploratory Analysis of Multidimensional Data: Types of Multidimensional Data Display. Remote measurements search. 5. Statistical testing of multidimensional random samples: Estimation of position and dispersion parameters. 6. Covariance analysis. Explanation of covariance matrix. Correlation matrix analysis. Types of correlation coefficients. 7. Analysis of main components of PCA. Importance of major components. PCA Graphics Utilities. PCA Diagnostics. 8. Factor Analysis FA. Factor analysis model and parameter estimation. Factor score estimation, factor rotation. 9. CCA correlation analysis. The nature of the method. Significance test of canonical correlations. 10. Discriminant Analysis DA: Classification of Objects. The nature of the method, the DA procedure, and the spooling rules. 11. Logistic regression LR. The nature of the method and the process of logistic regression. Parameter estimates. 12. Cluster analysis of CLU. The nature of cluster analysis. Similarity and distance measures. Suitability of data standardization. 13. Mapping objects by multidimensional scaling of MDS. The nature of the method and the process of multidimensional scaling.
|
|
Learning activities and teaching methods
|
unspecified, Monologic (reading, lecture, briefing)
- Home preparation for classes
- 120 hours per semester
|
|
Learning outcomes
|
Learning outcomes of the course unit The aim of the course is to acquaint students with modern methods of multidimensional data evaluation, conditions of use, selection of adequate method, and above all interpretation of results.
Student after passing the subject - demonstrates the knowledge of specific tools for displaying multidimensional data, finding outliers, internal data dependencies - can use available software (STATISTICA), can independently evaluate model examples
|
|
Prerequisites
|
Completion of the course NNSZD.
|
|
Assessment methods and criteria
|
Oral examination
Process independently assigned model cases. Create and evaluate one example from your own or retrieved data (literature, Internet).
|
|
Recommended literature
|
-
JAVŮREK, M., TAUFER, I. Vyhodnocování experimentálních dat. 2. vydání. Pardubice, 2018. ISBN 978-80-270-3611-0.
-
MELOUN, M., MILITKÝ, J., HILL, M. Počítačová analýza vícerozměrných dat v příkladech.. Praha, 2005. ISBN 80-200-1335-0.
|