Ανάλυση της τιμής του αργού πετρελαίου και εφαρμογή σύγχρονων μεθόδων για την πρόβλεψή της
View/
Keywords
Machine learning ; Timeseries forecasting ; ARIMA ; Crude oilAbstract
In the present work, crude oil is studied as a factor of influence of the world economy and as an investment and an empirical analysis is made of its price and the factors related to it.
The oil’s price initiation is a very important issue for the world community as it is bilaterally related to most of the financial indicators of the economy, which it affects to a great extent and is also influenced by them. According to many people, it is the main factor regulating the functioning of the world economy since it remains the main source of energy on the planet. Oil’s price fluctuations have a direct impact on all macroeconomic variables such as inflation, international interest rates and exchange rates, and an uncontrolled upward trend in the price of oil can lead to a slowdown in global economic activity.
Nowadays crude oil is an investment choice as a factor of portfolio diversification attracting more and more investors and making it imperative to understand the characteristics of its market. For this reason, a multitude of researchers are examining the factors that change the price of crude oil by developing predictive models based on historical records of the prices of other financial assets and commodities but mainly the price of crude oil itself. However, predicting the price of oil remains one of the most difficult problems in the world of analysis due to its high volatility.
The purpose of this work is to provide useful information to oil market stakeholders by studying global market movements and creating an appropriate predictive model comparing in terms of accuracy machine learning applications. In order to draw useful conclusions about oil price volatility, the correlations of gold, silver, 10-year T-Note, US inflation and the NASDAQ 100 monthly stock index will be analyzed on a monthly basis from January 2000 until December 2020. A descriptive analysis of the above indicators will be carried out first in order to detect possible correlation between them, followed by the creation of appropriate predictive machine learning models based on the historical records of the oil price.
Postgraduate Studies Programme
Πληροφοριακά Συστήματα και ΥπηρεσίεςDepartment
Σχολή Τεχνολογιών Πληροφορικής και Επικοινωνιών. Τμήμα Ψηφιακών ΣυστημάτωνNumber of pages
54Language
GreekCollections
Except where otherwise noted, this item's license is described as
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα