National Repository of Grey Literature 23 records found  1 - 10nextend  jump to record: Search took 0.00 seconds. 
Statistical machine learning with applications in music
Janásková, Eliška ; Večeř, Jan (advisor) ; Hlávka, Zdeněk (referee)
The aim of this thesis is to review the current state of machine learning in music composition and to train a computer on Beatles' songs using research project Magenta from the Google Brain Team to produce its own music. In order to explore the qualities of the generated music more thoroughly, we restrict our- selves to monophonic melodies only. We train three deep learning models with three different configurations (Basic, Lookback, and Attention) and compare generated results. Even though the generated music is not as interesting as the original Beatles, it is quite likable. According to our analysis based on musically informed metrics, generated melodies differ from the original ones especially in lengths of notes and in pitch differences between consecutive notes. Generated melodies tend to use shorter notes and higher pitch differences. In theoretical background, we cover the most commonly used machine learning algorithms, introduce neural networks and review related work of music generation. 1
Parameter choice in portfolio optimization problems based on out-of-sample performance
Vaňková, Kateřina ; Kopa, Miloš (advisor) ; Večeř, Jan (referee)
This thesis investigates three optimization models using the rolling window method. These models are based on maximizing profits and minimizing risk. Two statistics are considered in the models: expected value and a risk measure. Risk measures analyzed in this thesis are: the variance, the Conditional Value-at-Risk at a specified confidence level, and the Mean Absolute Deviation. Models are tested on the real US stock data of ten companies in the time period of twenty years: from January 30th, 1999 to January 30th, 2019. The aim of this thesis is to analyze these models using the rolling window method and to investigate its sensitivity towards changes in the values of several parameters in order to identify the best parameter setting.
Volatility modeling
Jurka, Vojtěch ; Prášková, Zuzana (advisor) ; Večeř, Jan (referee)
In the thesis we deal with modelling volatility conditional on past shocks. Traditional ARCH and GARCH models proposed by Engle(1982) and Bollerslev(1986) are investigated as well as several generalizations of GARCH model that capture asymmetric reaction on positive and negative excess returns, namely GJR-GARCH, TGARCH and EGARCH. Selected models are then applied to four commodities traded on Chicago Mercantile Exchange that represent various sectors of commodity market. Our first key finding is that in short horizon all considered models have similar performance, while in longer horizon, EGARCH and TGARCH give more precise results. The second is that, measured by an average percentage error, there is no significant difference in quality of predictions among selected assets across commodity sectors.
Reverse mortgage
Korotkov, Daniil ; Mazurová, Lucie (advisor) ; Večeř, Jan (referee)
ČSOB Pojišťovna, a. s., člen holdingu ČSOB Veřejné 1 / 1 20.7.2018 Abstract: At this moment, reverse mortgages are relatively new products on the Czech market and this thesis deals with their problematics. In this thesis, we describe the main risks related to reverse mortgages, namely, longevity risk and adverse evolution of property prices. Analyzing these risks we are modelling the underlying property prices, their future behavior, discount factors along with studying the risk models such as vector autoregression, hedonic model, repeat-sales and Wills-Sherris model. In practical part, we focus on estimating the parameters of Lee-Carter model and autoregression model of zero-coupon government bond as well as applying the results of the estimation to calculate various characteristics of reverse mortgages.
Market model with random inputs
Krch, Ivan ; Lachout, Petr (advisor) ; Večeř, Jan (referee)
The thesis deals with market models with random inputs represented by the newsvendor problem for which the randomness is given through a random number of customers. Presented work is divided into three chapters. In the first chapter we present the elementar newsvendor problem as stochastic programming problem with a fixed recourse. In the second chapter we present the multiplayer game theory adapted to the newsvendors problem. Moreover, in the second chapter we extend the problem by the second newsvendor on the market and in the third chapter we generalize the problem for n newsvendors on the market. We deal with the situations that arise in the chapters two and three from the game theory point of view and we study characteristics of a Nash equilibrium. Presented theory is demonstrated on illustrative examples in the ends of the two last chapters. 1
Optimal portfolios
Vacek, Lukáš ; Hurt, Jan (advisor) ; Večeř, Jan (referee)
In this diploma thesis, selected techniques for construction of optimal portfo- lios are presented. Risk measures and other criteria (Markowitz approach, Value at risk, Conditional value at risk, Mean absolute deviation, Spectral risk measure and Kelly criterion) are defined in the first part. We derived analytical solution for some cases of optimization problems, in some other cases there exists numeri- cal solution only however. Advantages and disadvantages, theoretical properties and practical aspects of software implementation in Wolfram Mathematica are also mentioned. Simulation methods suitable for portfolio optimization are brie- fly presented with their motivation in the second part. Multivariate distributions: normal, t-distribution and skewed t-distribution are presented in the third part with connection to optimization of portfolio with assumption of multivariate dis- tribution of financial losses. Optimization methods are illustrated on real data in the fourth part of this thesis. Analytical methods are compared with numerical ones. 1
Machine learning with applications to finance
Mešša, Samuel ; Hurt, Jan (advisor) ; Večeř, Jan (referee)
The impact of data driven, machine learning technologies across a wide variety of fields is undeniable. The financial industry, which relies heavily on predictive modeling being no exception. In this work we summarize two widely used machine learning models: support vector machines and neural networks, discuss their limitations and compare their performance to a more traditionally used method, namely logistic regression. Evaluation was done on two real world datasets, which were used to predict default of loan applicants and credit card holders formulated as a binary classification task. Neural networks and support vector machines either outperformed or showed comparable results to logistic regression with performance measured in receiver operator characteristic area under curve. In the second task neural networks outperformed both other models by a significant margin.
Alternative risk measures and their applications
Drobuliak, Matúš ; Hurt, Jan (advisor) ; Večeř, Jan (referee)
Title: Alternative risk measures and their applications Author: Matúš Drobuliak Department: Department of Probability and Mathematical Statistics Supervisor: Doc. RNDr. Jan Hurt, CSc., Department of Probability and Mathe- matical Statistics Abstract: The objective of this thesis is to discuss alternative measures of risk. We focused on the expectile value at risk, which we compared with conventional risk measures - namely value at risk and conditional value at risk. We also discussed its properties from the financial point of view. A numerical illustration is included in the thesis. Keywords: Value at risk, Conditional value at risk, Quantile, Expectile, Expectile value at risk iii
Favoritism Under Social Pressure: Evidence From English Premier League
Herrmann, Vojtěch ; Večeř, Jan (advisor) ; Hlávka, Zdeněk (referee)
The aim of this thesis is to study the extent to which the English Premier League referees are influenced by social pressure, especially by the home support and by the general popularity of the teams. Using regression analysis, we compare the actual length of the overtime, which is fully in the competence of the referee, with the predicted one from the usual game stoppages. Then we try to identify factors that contribute to any possible discrepancy. Our results suggest that the games tend to be extended beyond the expectations when the outcome of the game still can change, i.e., when the score differential at the time 90:00 is either zero or one. However, this extra extension happens almost regardless of the playing teams and thus we find no evidence for referee bias towards any specific team. However, a small bias towards the group of "Big" teams has been found, but only in the games in which the score differential was different from one.
Fixed-odds betting and real odds
Vojtík, Jakub ; Lachout, Petr (advisor) ; Večeř, Jan (referee)
The fixed-odds betting is currently very popular. Bookmakers set the odds on sport events, which represent an amount possible to win per each unit bet. The aim of this thesis is to study whether the odds can reflect the probabilities of the outcomes of sport matches, based on which then try to tell if the same can go for the bettors and their bets. First, there are derived several models for setting the odds. Then there is constructed a hypothesis test to test a validity of one of these models. This cannot be rejected, thanks to which there is estimated the dependence of the real probabilities on the odds. It turns out that with the exception of probabilities close to 0 and 1 the estimation might work. The conclusion of the thesis is the statement that odds could a be a good indicator of the probabilities and in such case also the bets of the bettors would correspond to them. 1

National Repository of Grey Literature : 23 records found   1 - 10nextend  jump to record:
Interested in being notified about new results for this query?
Subscribe to the RSS feed.