Graduate theses

Current Record: 5 of 14

[Add to Basket]
Identifier	000432644
	h
Title	Applying machine learning techniques for predictive maintenance of professional equipment
Alternative Title	Εφαρμογές τεχνικών μηχανικής μάθησης για την προληπτική συντήρηση επαγγελματικού εξοπλισμού
Author	Σφακιανάκης, Παντελεήμων
Thesis advisor	Σμαραγδάκης, Κώστας Ταρουδάκης, Μιχαήλ
Abstract	Predictive maintenance is a concept becoming increasingly popular amongst businesses thanks to the huge economic benefits it provides to them. In this thesis the issue of predictive failure incidents for a water pump will be discussed. Three different Recurent Neural Networks will be compared in order to realise which is more appropriate for this task. These neural networks are: Simple RNN, Long Short Term Memory and Gated Recurrent Unit. Simple RNN uses less controlling knobs than the GRU which uses less than the LSTM.
Language	English, Greek
Subject	Προγνωστική συντήρηση
Issue date	2020
Collection	School/Department--School of Sciences and Engineering--Department of Mathematics and Applied Mathematics--Graduate theses
	Type of Work--Graduate theses
Permanent Link	https://elocus.lib.uoc.gr//dlib/5/9/7/metadata-dlib-1601540475-703978-991.tkl
Views	215

Digital Documents
	Download document View document Views : 26