|
|
|
||
|
An introduction to traditional and modern methods of multivariate statistics.
Last update: Omelka Marek, doc. Ing., Ph.D. (01.06.2023)
|
|
||
|
The aim is ťo teach students basic approaches of statistical learning and how they relate to classical methodology of multivariate statistical analysis. Last update: Mizera Ivan, prof. RNDr., CSc. (20.09.2026)
|
|
||
|
Requirements for obtaining the credit (zápočet): participation in the exercises (max 3 absences) and continual solving of the assigned problems (acquiring at least 36 credits, where one solved problem typically amounts to one credit). The nature of these requirements precludes any possibility of additional attempts to obtain the class credit. Acquired credit is a condition for attending the examination, which will be in the written form, and apart from simple questions similar to those covered in the exercises will contain also questions regarding principles, motivations, algorithms, and applications of the techniques covered in the lectures. Last update: Mizera Ivan, prof. RNDr., CSc. (20.09.2026)
|
|
||
|
Venables W.N. Ripley B.D.: Modern Applied Statistics with S, 4th edition, Springer, New York, 2002. Mardia K.V., Kent J.T., Bibby J.M.: Multivariate Analysis. Academia Press. London, 1979. Rao C.R.: Linear Statistical Inference and Its Applications. 2nd edition. Wiley. New York, 1973. (there is a Czech translation) Hastie, T., Tibshirani, R., Friedman, J.: The Elements of Statistical Learning: Data Mining, Inference, and Prediction, 2nd edition, Springer, New York, 2009. Izenman, A.J. Modern Multivariate Statistical Techniques: Regression, Classification, and Manifold Learning, Springer, New York, 2008. Last update: Mizera Ivan, prof. RNDr., CSc. (20.09.2026)
|
|
||
|
Lectures, exercises, discussions. Last update: Mizera Ivan, prof. RNDr., CSc. (20.09.2026)
|
|
||
|
The exam will be in the written form, consisting of several questions related to the syllabus, as was presented in the lectures and/or exercises. The emphasis will be on correct understanding of the statistical (that is, data-analytical) substance and the interpretation of the covered methods. Last update: Mizera Ivan, prof. RNDr., CSc. (20.09.2026)
|
|
||
|
Theory of predictive classification and principles of some of its specific methods: classification via density estimation, flexible regression methods like (regularized) logistic regression, classification and regression trees, artificial neural networks,. Methods based on multivariate normal distribution (discriminant analysis} and ich connection to classical multivariate methods like MANOVA, and its traditional special cases connected with the names of Hotelling and Wishart. Dimension reduction methods: principal components, factor analysi, canonical correlations, correspondence analysis. Data-analytic methods utilizing distances: multidimensional scaling, cluster analysis. Last update: Mizera Ivan, prof. RNDr., CSc. (20.09.2026)
|
|
||
|
Fluency in linear algebra, theory of probability, and mathematical statistics. Last update: Mizera Ivan, prof. RNDr., CSc. (20.09.2026)
|