Σχεδίαση και ανάπτυξη κατηγοριοποιητή Bayesian network στο κατανεμημένο περιβάλλον του Apache Spark
Loading...
Date
2023-06-15
Authors
Ακαρέπης, Ιωάννης
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Every year the volume of information is growing at a high rate. Existing computer systems are reaching the
limits of handling such data, and as a result, a more modern approach is required in the field of data science
to deal efficiently with the problems of the new era. Distributed systems, offer such a modern approach,
and as such, more and more machine learning models for automatic decision-making are being transfered
to distributed environments such as that of Apache Spark. In the work presented in this thesis, a new
classification model based on Bayesian networks is proposed, which utilizes the distributed environment
of Apache Spark. This model is experimentally evaluated with corresponding classical Bayesian Network
models as well as different classifiers, in order to show that this model can be competitive and complete in
this new era of information.
Description
Keywords
Μηχανική μάθηση, Κατανεμημένα συστήματα