Συνεργατικό φιλτράρισμα βασισμένο στους χρήστες με την βοήθεια γενετικών αλγορίθμων
User-based collaborative filtering with the help of genetic algorithms

View/ Open
Keywords
Συστήματα συστάσεων ; Συνεργατικό φιλτράρισμα ; Γενετικοί αλγόριθμοι ; Τεχνητή νοημοσύνη ; Παραγοντοποίηση πινάκων ; Ημί-εποπτευόμενη μάθηση ; Χαμηλόβαθμες υπερπαραμέτροι πινάκωνAbstract
Recommender systems are an important subfield of Artificial
Intelligence, making use of varied techniques, in order to come up with
useful recommendations for the users of the RS. Genetic Algorithms
are a category of algorithms used in order to solve Artificial Intelligence
problems, where the search space for the solution is large and hard to
search.In this undergraduate thesis two such algorithms where
created, GenRec1 and GenRec0. They search for optimal connections
between users, helping predict the rating these users give on certain
items. The two models are analyzed on the basis of their methodology,
their implementation in terms of code, as well as their results, in
comparison to other common methods in the field.