Αποδοτικές μέθοδοι εκτέλεσης παραγοντικών σχεδιασμών σε ομάδες (blocks)
Efficient blocking of factorial designs
View/
Keywords
Παραγοντικοί σχεδιασμοί σε ομάδες ; blockingAbstract
In the present paper we study factorial designs with two levels, which are found in several experiments as well as in industry. In the first chapter we analyse full factorial designs and specifically the full 2k factorial design, as well as the fractional 1/2k factorial designs. The main objective of the work is the orthogonal arrays of two levels and some "good" properties they have. In the second chapter we will apply the technique of blocking, with the aim of optimal performance of the design. This technique is a way to overcome the frequent problem of inhomogeneous conditions that occurs during the various runs of the experiment. In the literature we find several references to optimization techniques of blocked factorial designs and orthogonal formations, such as by Sun, Cheng and Wu, Das and others.
In the third chapter we will use the well-known criterion of D-efficiency to measure the efficiency of blocked designs, in which the effects of the blocked factors are uncorrelated with the main effects. In addition, with this criterion the parameters of the linear regression model consisting of the mean, the block factor, the main effects, and the 2nd order interactions, are calculated with the greatest possible efficiency.
Finally, in the fourth chapter we will present designs with 4, 5 and 6 columns and we will choose as the optimal grouping option the column that shows the highest value of the above criterion.
Postgraduate Studies Programme
Εφαρμοσμένη ΣτατιστικήDepartment
Σχολή Χρηματοοικονομικής και Στατιστικής. Τμήμα Στατιστικής και Ασφαλιστικής ΕπιστήμηςNumber of pages
54Language
GreekCollections
Except where otherwise noted, this item's license is described as
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα