E-Locus - Institutional Repository of the University of Crete - Ομαδοποίηση "εν πτήσει" Δοσοληψιών σε Ομάδες με Παρόμοια Χαρακτηριστικά Φόρτου Εργασίας

Home Ομαδοποίηση "εν πτήσει" Δοσοληψιών σε Ομάδες με Παρόμοια Χαρακτηριστικά Φόρτου Εργασίας

Results - Details

[Add to Basket]

Identifier

uch.csd.msc//1997kloukinas

Title

Ομαδοποίηση "εν πτήσει" Δοσοληψιών σε Ομάδες με Παρόμοια Χαρακτηριστικά Φόρτου Εργασίας

Alternative Title

On-the-fly clustering of OLTP workload characterizations

Creator

Kloukinas, Christos

Contributor

Χ. Νικολάου

Abstract

The following work deals with clustering of transactions in a distributed OLTP system, according to their workload characteristics. It continues the work done in \cite{Labrinides}, differentiating from it by the fact that it no longer deals with static sets of data and thus with batch clustering algorithms. Instead, it deals with a dynamic sequence of data which calls upon the use of an ``on the fly'' clustering algorithm. For this purpose, two dynamic clustering algorithms have been implemented. The first one is based on an artificial neural network of the adaptive kmeans family of networks, while the second one is based on a graph - theoretical clustering method.

Subject

α) Παράλληλα και Κατανεμημένα Συστήματα, β) Μηχανική Οραση και Ρομποτική.

Issue date

1997-07-01

Date available

1997-10-2

Collection

School/Department--School of Sciences and Engineering--Department of Computer Science--Post-graduate theses

Type of Work--Post-graduate theses

Views

432

Digital Documents
	Download document View document Views : 3
	View document
	View document