Syllabus
1. Basic probability and statistics - random events and variables, normal distribution, random sample, parameter estimating, testing of statistical hypotheses.
2. Linear and logistic regression - description of a continuous or a binary variable using a set of regressors.
3. Testing hypotheses concerning the mean - t-tests and their non-parametric alternatives.
4. Analysis of variance - one-way and two-way ANOVA, hierarchical model, latin squares.
5. Factorial experimental design - two-level factorial design, fractional factorial design.
6. Principal component analysis - interpretation of the principal components in data sets.
7. Cluster analysis - hierarchical and non-hierarchical clustering algorithms.
8. Discriminant analysis - classification of objects to a given set of groups.
9. Time series analysis - modelling and forecasting.
Annotation
The course is dedicated to some methods of data analysis which enable to obtain as much as possible information from large amounts of measurements in medicine, pharmaceutical or environmental research etc. The aim is to show the practical use of the methods, therefore the analysis of real data on computer using an appropriate statistical software is an important part of the course.
Two types of problems are solved: investigating the influence of several factors on a variable by means of experimental design and classification of a large data set by means of multivariate statistical analysis. Regression and time series models are also explained and applied to data.