Multivariate Analysis

Most popular methods of multivariate data analysis are presented in the course. They can be used to analyse complex phenomena characterized by many variables. The course covers graphic presentations of multivariate data, applied multivariate regression, ordering of multidimensional objects, cluster analysis, basics of Factor Ananlysis and Principal Component Analysis and Multidimentional Scaling – all of them performed in Statistica Software.

Who should attend?

• For those who are interested in the analysis of phenomena described by many variables with application in medicine, social sciences, economics, market research, psychology and many other fields

What will you learn?

• How to use the most popular techniques of multivariate data analysis
• How to extract useful knowledge from your data
• How to graphically present multivariate data
• How to select the best subset of variables and build your own model
• How to look for relations between variables
• How to build rankings of objects described by many variables
• How to reduce dimensiality of your problem
• How to find groups of similar objects
• How to be more fluent in Statistica

Training programme:

1. Introduction
   • Measurement scales
   • Multivariate random variable. Finite mixture distribution.
   • Statistical hypotheses testing
2. Graphical presentation of multivariate data
3. Applied multiple regression
   • Linear multiple regression model
   • Essential steps in applied regression analysis
   • Stepwise regression
   • Goodness-of-fit measures
   • Analysis of residuals
4. Linear ordering of multidimentional objects (rankings)
   • Identification and selection od diagnostic variables
   • Normalization and standarization
   • Aggregation formulas
5. Cluster analysis
   • Tasks and problems in cluster analysis
   • Distance measures
   • Agglomerative methods
   • K-means method
   • Comparing partitions
   • Dynamic cluster analysis
6. Multidimentional scaling
7. Factor analysis and Principal Component Analysis
   • General ideas and aims of FA and PCA
   • Interpretation of eigenvalues and factor loadings
   • Identification of factors

Training hours: 8:15 AM – 3:30 PM

The training includes:

• comprehensive training materials
• certificate of completion
• coffee breaks and lunch (for on-site training)

Training types:

• online or on-site
• the format is indicated with the training date

For online trainings:

• live sessions via MS Teams
• real-time interaction with the instructor

Prerequisites:

• computer skills in a Windows environment

Price: 930 EUR net

Price for academic Statistica users with an active StatSoft Polska support agreement (-15%)*: 791 EUR net