LF:DSAN02 Advanced Data Analysis for Neu - Course Information
DSAN02 Advanced Data Analysis for NeuroscienceFaculty of Medicine
- Extent and Intensity
- 0/0. 5 credit(s). Type of Completion: k (colloquium).
- prof. RNDr. Ladislav Dušek, Ph.D. (lecturer)
RNDr. Eva Koriťáková, Ph.D. (lecturer)
RNDr. Simona Littnerová, Ph.D. (seminar tutor)
Mgr. et Mgr. Petr Dluhoš (seminar tutor)
- Guaranteed by
- prof. RNDr. Ladislav Dušek, Ph.D.
Institute of Biostatistics and Analyses - Other Departments for Educational and Scientific Research Activities - Faculty of Medicine
Contact Person: prof. RNDr. Ladislav Dušek, Ph.D.
Supplier department: Institute of Biostatistics and Analyses - Other Departments for Educational and Scientific Research Activities - Faculty of Medicine
- DSAN01 Data analysis for Neuroscience
Basic knowledge of biostatistics and data analysis. It is recommended to attend DSAN01 Data analysis for Neuroscience beforehand.
- Course Enrolment Limitations
- The course is only offered to the students of the study fields the course is directly associated with.
The capacity limit for the course is 20 student(s).
Current registration and enrolment status: enrolled: 4/20, only registered: 0/20
- fields of study / plans the course is directly associated with
- there are 16 fields of study the course is directly associated with, display
- Course objectives
- The course objectives are to improve knowledge and practical skills of data analysis of the students by teaching them advanced multivariate methods of medical data analysis with respect to particularities of large data files in the neuroscience research. The main emphasis will be laid on correct application of the methods and interpretation of results. Theory will be followed by practical demonstrations in the software SPSS, R, and MATLAB which are freely available at Masaryk University.
- Learning outcomes
- After the course, the students will be able to:
• Prepare data sets for multivariate analysis
• Describe and visualize multivariate data
• Choose appropriate distance or similarity metrics
• Select and apply relevant clustering method
• Reduce and transform multivariate data using ordination methods
• Classify data using various methods for discriminant analysis
• Evaluate classification accuracy
• Interpret results obtained by multivariate analyses
- 1. Introduction to multivariate data analysis: Importance and objectives of multivariate data analysis. Examples of multivariate data analyses. Data matrices. Description and visualization of multivariate data using tables and graphs.
- 2. Multivariate statistical tests and distributions: Multivariate descriptive statistics – mean vector, covariance matrix, correlation matrix. Multivariate normal distribution. Multivariate t-test. Multivariate analysis of variance. Multivariate data transformations.
- 3. Distance and similarity metrics in multidimensional space: Distance metrics – e.g. Euclid, Hamming, Mahalanobis distances. Similarity metrics – e.g. Jaccard, Tanimoto, Sokal-Michener indices. Association matrix.
- 4. Cluster analysis: Basics and objectives of clustering analysis. Hierarchical cluster analysis – agglomerative methods (e.g. nearest neighbor clustering, furthest neighbor clustering, average linkage clustering, centroid method, Ward‘s method), divisive methods. Non-hierarchical cluster analysis (e.g. k-means clustering, x-means clustering, partitioning around medoids). Identification of optimal number of clusters.
- 5. Ordination methods I: Principles of data reduction. Selection and extraction of variables. Principal component analysis (PCA). Factor analysis (FA).
- 6. Ordination methods II: Independent component analysis (ICA). Overview of other ordination methods – correspondence analysis (CA), multidimensional scaling (MDS), redundancy analysis (RDA), canonical correlation analysis (CCorA).
- 7. Classification I: Principles and objectives of classification. Discriminant analysis based on discriminant functions – Bayes classifier. Discriminant analysis based on minimal distance. Discriminant analysis based on boundaries – Fisher’s linear discriminant analysis (LDA). Association of discriminant analysis with logistic regression.
- 8. Classification II: Discriminant analysis based on boundaries – support vector machines (SVM). Overview of other classification methods – decision trees and random forests, neuronal networks. Evaluation of classification accuracy – cross-validation, comparison of classification results with random classification, comparison of classification results of two or more classifiers.
- • HEBÁK, Petr. Vícerozměrné statistické metody (1). Informatorium, Praha. 2004, 239 s., ISBN 8073330253
- • DUDA R. O., HART P. E., STORK D. G., 2000: Pattern Classification. Wiley-Interscience, New York, 680 pp., ISBN 0471056693
- • BISHOP C., 2006: Pattern Recognition and Machine Learning. Springer, New York, 738 pp., ISBN 0387310738
- • THEODORIDIS, S. et al., 2010: Introduction to pattern recognition: a MATLAB approach. Academic Press, Amsterdam, 219 pp., ISBN 9780123744869
- • JOHNSON, R. et al., 2007: Applied multivariate statistical analysis. 6th ed. Prentice Hall, Upper Saddle River, N.J., 773 pp., ISBN 9780135143506
- Teaching methods
- Teaching is interactive and based on solving real problems and examples using advanced multivariate methods. The examples and study materials will be available beforehand. Students will be allowed to prepare examples of problems in their analyses (PhD theses, research activities etc.). These problems will be discussed and solved in following lectures.
- Assessment methods
- Course is finished by colloquium, consisting of analyses of sample data files using computer which test knowledge and skills of data analysis using advanced multivariate methods.
- Language of instruction
- Further comments (probably available only in Czech)
- Study Materials
The course is taught annually.
The course is taught: in blocks.
Information on the extent and intensity of the course: 8 x 3 hod.