magistrsko delo
Andrej Sevčnikar (Author), Tatjana Welzer-Družovec (Mentor)

Abstract

V magistrskem delu smo se osredotočili na napovedovanje značilnosti prometa naosnovi podatkov časovnih vrst. Za dosego cilja smo v magistrski nalogi proučili področje integracije podatkov in podatkovnih skladišč, z uporabo katerega smo izdelali celovit pogled na podatke, ki je nujno potreben za nadaljnjo manipulacijo s podatki. Osrednja tematika magistrske naloge so modeli za napovedovanje podatkov časovnih vrst, s katerimi smo napovedovali gostoto prometa, pri tem pa preverjali napovedno napako. Določili smo najbolj optimalne modele za napovedovanje gostote prometa. Pokazali smo, da je pravilna predpriprava podatkov oziroma učenje modelov na podlagi dobro zasnovanih časovnih okvirjev izredno pomembno za pravilno napoved prometa, hkrati pa določili optimalno število časovnih okvirjev, pri katerih je napovedna napaka še sprejemljiva.

Keywords

podatkovno skladišče;integracija podatkov;časovne vrste;podatkovno rudarjenje;gručenje;promet;napovedovanje;

Data

Language: Slovenian
Year of publishing:
Typology: 2.09 - Master's Thesis
Organization: UM FERI - Faculty of Electrical Engineering and Computer Science
Publisher: [A. Sevčnikar]
UDC: 004.62(043)
COBISS: 17414678 Link will open in a new window
Views: 1608
Downloads: 154
Average score: 0 (0 votes)
Metadata: JSON JSON-RDF JSON-LD TURTLE N-TRIPLES XML RDFA MICRODATA DC-XML DC-RDF RDF

Other data

Secondary language: English
Secondary title: INOFMRATION SUPPORT FOR TRAFFIC CHARACTERISTIC PREDICTION BASED ON TIME SERIES DATA
Secondary abstract: In the master thesis we focus on the prediction of traffic characteristics based on time series data. To achieve this goal we investigate areas of data integration and data warehousing in order to attain a comprehensive view of data which is vital for further data manipulation. The main area of this thesis are models for the prediction of time series data which make it possible to predict traffic whilst checking the predictive error. We determinethe optimal model for predicting the density of traffic. We show thatthe learning models based on well-designed time frames are very important for precise traffic prediction. Additionally, we define the optimal number of time frames in which the prediction error is acceptable.
Secondary keywords: data warehouse;data integration;time series;data mining;clustering;traffic;forecasting;
URN: URN:SI:UM:
Type (COBISS): Master's thesis
Thesis comment: Univ. v Mariboru, Fak. za elektrotehniko, računalništvo in informatiko
Pages: XI, 108 f.
ID: 8728154