SKOPAL, M. Analýza a předpověď ekonomických časových řad pomocí vybraných statistických metod [online]. Brno: Vysoké učení technické v Brně. Fakulta strojního inženýrství. 2019.
The thesis of Mr. Martin Skopal aim to analyzed and forecasted the economic time series using selected statistical methods. The thesis describes the fundaments of time series, their predictions and create original method, which combines the different statistical approaches using weigh functions. Part of the work dealing with model testing on selected real time series of stock market. The student fulfilled all the objectives of the thesis according to the assignment, worked independently, regularly consulted and actively searched for literal sources of information. Student has shown great initiative in combining a large number of statistical models as well as in the selection of the tested series. The created model in MATLAB is able to process any time series automatically with its architecture and find the most appropriate combination of statistical models. From a formal point of view, the thesis is well written except for a few things like the order of the cited literature or the absence of a list of abbreviations. The thesis is written in high mathematical level and meets the requirements for the graphic form.
| Kritérium | Známka | Body | Slovní hodnocení |
|---|---|---|---|
| Splnění požadavků a cílů zadání | A | ||
| Postup a rozsah řešení, adekvátnost použitých metod | A | ||
| Vlastní přínos a originalita | B | ||
| Schopnost interpretovat dosažené výsledky a vyvozovat z nich závěry | B | ||
| Využitelnost výsledků v praxi nebo teorii | B | ||
| Logické uspořádání práce a formální náležitosti | B | ||
| Grafická, stylistická úprava a pravopis | A | ||
| Práce s literaturou včetně citací | B | ||
| Samostatnost studenta při zpracování tématu | A |
Master’s thesis of Martin Skopal deals with the development of an automated algorithm for the analysis and forecasting of financial time series. The objective was to formulate an algorithm that would analyze and forecast financial time series without the need of human intervention. At the beginning of his thesis, the author introduces basic characteristics of time series such as stationarity, heteroscedasticity, seasonality, etc. Then a brief description of several time series analysis and forecasting models and methodologies is presented. The proposed forecasting algorithm, which consists of several procedures and models, was implemented in MATLAB. The algorithm was tested on several financial time series, such as the Dow Jones Industrial Average index or the Coca-Cola Company stock price. The author concluded that the proposed algorithm generally outperformed Box-Jenkins and ETS models in the conducted tests. The master’s thesis of Martin Skopal is rather well-written and relatively easy to follow. The author uses flow charts to graphically illustrate the working of the proposed algorithm. Martin Skopal has shown a rather deep understanding of the methods and models used in the analysis and forecasting of time series. He has also demonstrated his skills in the implementation of the proposed algorithm in MATLAB.
| Kritérium | Známka | Body | Slovní hodnocení |
|---|---|---|---|
| Splnění požadavků a cílů zadání | A | ||
| Postup a rozsah řešení, adekvátnost použitých metod | A | ||
| Vlastní přínos a originalita | A | ||
| Schopnost interpretovat dosaž. výsledky a vyvozovat z nich závěry | A | ||
| Využitelnost výsledků v praxi nebo teorii | A | ||
| Logické uspořádání práce a formální náležitosti | A | ||
| Grafická, stylistická úprava a pravopis | A | ||
| Práce s literaturou včetně citací | A |
eVSKP id 116396