Εφαρμογή τεχνικών πολυμεταβλητής ανάλυσης σε ιατρικά και κοινωνικο-οικονομικά προβλήματα
Application of multivariate analysis techniques in medical, financial and social problems
![Thumbnail](/xmlui/bitstream/handle/unipi/13818/Skyfa_18021.pdf.jpg?sequence=6&isAllowed=y)
View/ Open
Keywords
Τεχνικές ; Πολυμεταβλητή ανάλυση ; Κύριες συνιστώσες ; Ανάλυση κύριων συνιστωσών ; Ανάλυση παραγόντωνAbstract
Multivariate Analysis is a subdivision of Statistics encompassing a large amount of methods which aim to collect, study and analyze data consisting of multiple variables’ measurements in a set of experimental units.
The most commonly used techniques of Multivariate Statistical Analysis are:
• Principal Component Analysis,
• Factor Analysis,
• Cluster Analysis,
• Discriminant Analysis,
• Correspondence Analysis.
Every researcher can choose to apply one or more of the above techniques in order to reduce the number of variables used without loosing valuable information, group data into similar subsets, investigate the existence of dependence between variables, create forecast models and carry out hypothesis testing for parameters related to the studied data.
The purpose of this thesis is the presentation of the first two methods and the application of them in a case study, concerning medical data, in order to extract useful conclusions.