Matrix factorization techniques for recommender systems

In this thesis we study two basic matrix factorization techniques used in recommender systems, namely batch and stochastic gradient descent. Furthermore, data from Epinions.com, consisting of 40163 users and 139738 items is studied and statistically analyzed into its characteristic classes (i.e. use...

Πλήρης περιγραφή

Αποθηκεύτηκε σε:

Λεπτομέρειες βιβλιογραφικής εγγραφής
Κύριοι συγγραφείς:	Mavridis, Andreas, Μαυρίδης, Ανδρέας
Άλλοι συγγραφείς:	Ampazis, Nicholas
Γλώσσα:	en_US
Δημοσίευση:	2018
Θέματα:	Gradient Descent Matrix Factorization Epinions Recommender Systems Συστήματα υποδείξεων Παραγοντοποίηση πινάκων Recommender systems (Information filtering) (URL: http://id.loc.gov/authorities/subjects/sh2007003098) Factorization method (Quantum theory) (URL: http://id.loc.gov/authorities/subjects/sh2007003571) Gradient
Διαθέσιμο Online:	http://hdl.handle.net/11610/18038
Ετικέτες:	Προσθήκη ετικέτας Δεν υπάρχουν, Καταχωρήστε ετικέτα πρώτοι!

_version_	1828461178053984256
author	Mavridis, Andreas Μαυρίδης, Ανδρέας
author2	Ampazis, Nicholas
author_facet	Ampazis, Nicholas Mavridis, Andreas Μαυρίδης, Ανδρέας
author_sort	Mavridis, Andreas
collection	DSpace
description	In this thesis we study two basic matrix factorization techniques used in recommender systems, namely batch and stochastic gradient descent. Furthermore, data from Epinions.com, consisting of 40163 users and 139738 items is studied and statistically analyzed into its characteristic classes (i.e. users with only high ratings, items with only low ratings e.t.c. ). Matrix factorization performance is also examined and prediction accuracy is associated with user and item classes.
id	oai:hellanicus.lib.aegean.gr:11610-18038
institution	Hellanicus
language	en_US
publishDate	2018
record_format	dspace
spelling	oai:hellanicus.lib.aegean.gr:11610-180382025-03-14T11:56:01Z Matrix factorization techniques for recommender systems Αλγόριθμοι Παραγοντοποίησης Πινάκων σε Συστήματα Υποδείξεων Mavridis, Andreas Μαυρίδης, Ανδρέας Ampazis, Nicholas Αμπαζής, Νικόλαος Gradient Descent Matrix Factorization Epinions Recommender Systems Συστήματα υποδείξεων Παραγοντοποίηση πινάκων Recommender systems (Information filtering) (URL: http://id.loc.gov/authorities/subjects/sh2007003098) Factorization method (Quantum theory) (URL: http://id.loc.gov/authorities/subjects/sh2007003571) Gradient In this thesis we study two basic matrix factorization techniques used in recommender systems, namely batch and stochastic gradient descent. Furthermore, data from Epinions.com, consisting of 40163 users and 139738 items is studied and statistically analyzed into its characteristic classes (i.e. users with only high ratings, items with only low ratings e.t.c. ). Matrix factorization performance is also examined and prediction accuracy is associated with user and item classes. Η παρούσα διπλωματική εργασία, πραγματεύεται δύο βασικούς αλγόριθμους παραγοντοποίησης πινάκων που χρησιμοποιούνται στα συστήματα υποδείξεων. Παράλληλα, δεδομένα από την ιστοσελίδα Epinions.com για 40163 χρήστες και 139738 αντικείμενα, μελετώνται και αναλύονται στατιστικά σε χαρακτηριστικές κλάσεις. Εξετάζεται η απόδοση των παραγοντοποιήσεων των πινάκων και τα αποτελέσματα των προβλέψεων συσχετίζονται με τις κλάσεις των χρηστών και των αντικειμένων. 2018-02-28T13:17:53Z 2018-02-28T13:17:53Z 2017-10 http://hdl.handle.net/11610/18038 en_US CC0 1.0 Παγκόσμια http://creativecommons.org/publicdomain/zero/1.0/ 77 σ. application/pdf Χίος
spellingShingle	Gradient Descent Matrix Factorization Epinions Recommender Systems Συστήματα υποδείξεων Παραγοντοποίηση πινάκων Recommender systems (Information filtering) (URL: http://id.loc.gov/authorities/subjects/sh2007003098) Factorization method (Quantum theory) (URL: http://id.loc.gov/authorities/subjects/sh2007003571) Gradient Mavridis, Andreas Μαυρίδης, Ανδρέας Matrix factorization techniques for recommender systems
title	Matrix factorization techniques for recommender systems
title_full	Matrix factorization techniques for recommender systems
title_fullStr	Matrix factorization techniques for recommender systems
title_full_unstemmed	Matrix factorization techniques for recommender systems
title_short	Matrix factorization techniques for recommender systems
title_sort	matrix factorization techniques for recommender systems
topic	Gradient Descent Matrix Factorization Epinions Recommender Systems Συστήματα υποδείξεων Παραγοντοποίηση πινάκων Recommender systems (Information filtering) (URL: http://id.loc.gov/authorities/subjects/sh2007003098) Factorization method (Quantum theory) (URL: http://id.loc.gov/authorities/subjects/sh2007003571) Gradient
url	http://hdl.handle.net/11610/18038
work_keys_str_mv	AT mavridisandreas matrixfactorizationtechniquesforrecommendersystems AT mauridēsandreas matrixfactorizationtechniquesforrecommendersystems AT mavridisandreas algorithmoiparagontopoiēsēspinakōnsesystēmataypodeixeōn AT mauridēsandreas algorithmoiparagontopoiēsēspinakōnsesystēmataypodeixeōn

Matrix factorization techniques for recommender systems

Παρόμοια τεκμήρια