Wyniki wyszukiwania - Biblioteka Nauki

Nowa wersja platformy, zawierająca wyłącznie zasoby pełnotekstowe, jest już dostępna.
Przejdź na https://bibliotekanauki.pl

Ograniczanie wyników

2 Studies in Logic, Grammar and Rhetoric

2 2010

2 Cwiklinska-Jurkowska M.

Znaleziono wyników: 2

Liczba wyników na stronie

Wyniki wyszukiwania

Wyszukiwano:
w słowach kluczowych: EXPLORATORY DATA ANALYSIS

Sortuj według:

Ogranicz wyniki do:

EXPLORATORY DATA ANALYSIS FOR THE HEMATOLOGICAL FEATURES. PART I. METHODOLOGY

100%

Cwiklinska-Jurkowska M.

Studies in Logic, Grammar and Rhetoric

2010

nr 21(34)

47-59

Part I of the work characterizes the data of human blood parameters. It describes examined biomedical data set and used multivariate statistical exploratory methods like PCA, FA, MDS and clusters analysis. To the factor analysis methods belongs the biplots visualization method. Biplots are simply the scatterplots with the superimposition of the variables. The work compares different other alternative multivariate exploratory data analysis procedures from methodological point of view. The described multivariate ordination methods are applied in the Part II of the work 'Application'.

EXPLORATORY DATA ANALYSIS FOR THE HEMATOLOGICAL FEATURES. PART II. APPLICATION

100%

Cwiklinska-Jurkowska M.

Studies in Logic, Grammar and Rhetoric

2010

nr 21(34)

61-80

Part II contains numerical and graphical exploratory analysis results and elements of their medical interpretation for the set of the hematological observations and hematological variables. On the basis of correlation matices, matrix of scatterplots with overlaid regression lines and two and three-dimensional biplots relationships between parameters of blood during hemoperfusion is examined. For comparison purposes also MDS and one-way and two-way cluster analysis are performed. Usefulness of applied methods of multivariate data ordination to inspect, e.g. variables' interdependencies was assessed. Applied methods gave very close results and the medical interpretation of the results confirm some physiological clotting ideas. The practical results confirm some hypothesis describing polymer-blood interactions. Additionally, the results of principal factor analysis and multidimensional metric scaling with cluster analysis are concordant. The variety of applied exploration data methods confirm results and give the possibility of looking at data from different point of views.