Analysing Country-Specific GDP Growth in Europe with Additive Models : master's thesis

Pršlja, Katarina (avtor), Friedl, Herwig (mentor), Košir, Tomaž (mentor)

In this Master's thesis, we concentrate on regression models, where the final goal is to build an additive model. Further, the purpose of this model is to examine the relationship between the response variable GDP and relevant indicators having the influence on GDP. However, we first introduce linear models, since they are a theoretical foundation for constructing additive models. They are flexible because they do not impose a linear relationship between variables but rather describe it by some smooth function derived from the data. Later in the thesis, we try to make a forecast for the year 2016 based on some established model, because only data up to 2016 were available at the time of work creation.

• Jezik: Angleški jezik
• Ključne besede: regression, parametric models, additive models, estimator, smooth functions, interval, testing, matrix
• Vrsta gradiva: Magistrsko delo/naloga
• Organizacija: FMF - Fakulteta za matematiko in fiziko
• Leto izida: 2019
• PID: 20.500.12556/RUL-106508
• UDK: 519.2
• COBISS.SI-ID: 18583129
• Datum objave v RUL: 28.02.2019

Sekundarni jezik

• Jezik: Slovenski jezik
• Naslov: Analiziranje BDP rasti v posameznih državah Evrope z aditivnimi modeli : magistrsko delo

Izvleček

V magistrskem delu se osredotočamo na obravnavo regresijskih modelov. Naš cilj je predstaviti razred aditivnih modelov, s katerim želimo preučiti razmerje med odvisno spremenljivko BDP in ustreznimi ekonomskimi kazalniki, ki imajo vpliv na BDP. Vendar pa se najprej seznanimo z linearnimi modeli, saj so teoretična osnova za konstrukcijo aditivnih modelov. Aditivni modeli so uporabni tudi zato, ker ne predpostavijo linearnega obnašanja med spremenljivkami, temveč jih opišejo z gladkimi funkcijami, ki so pridobljene na podlagi podatkov. Na koncu poskušamo napovedati BDP za leto 2016, ki temelji na uveljavljenem modelu, ker so v času nastajanja dela bili na voljo samo podatki do vključno 2016.

• Ključne besede: regresija, parametrični modeli, aditivni modeli, cenilka, gladke funkcije, interval, testiranje, matrika