|
|
Travelling Salesman Problem
Kolář, Adam ; Rozman, Jaroslav (referee) ; Zbořil, František (advisor)
The aim of this bachelor's thesis is to design a testing environment for the traveling salesman problem and compare the effectiveness of different approaches to the solution. The first part discussed the possibility of genetic algorithms, depending on the setting of a crossover, mutations and population size. In the second part, there is the same problem using two types of neural networks. The representative of the self-learning net was chosen Kohonen neural network. Hopfield neural network represents a method of minimizing the energy function with fixed coefficients. At both neural networks, there were described possible advantages and disadvantages. In the end, all the findings were interpreted in a global context.
|
|
|
Coevolution of Cartesian Genetic Algorithms and Neural Networks
Kolář, Adam ; Král, Jiří (referee) ; Zbořil, František (advisor)
The aim of the thesis is to verify synergy of genetic programming and neural networks. Solution is provided by set of experiments with implemented library built upon benchmark tasks. I've done experiments with directly and also indirectly encoded neural netwrok. I focused on finding robust solutions and the best calculation of configurations, overfitting detection and advanced stimulations of solution with fitness function. Generally better solutions were found using lower values of parameters n_c and n_r. These solutions tended less to be overfitted. I was able to evolve neurocontroller eliminating oscilations in pole balancing problem. In cancer detection problem, precision of provided solution was over 98%, which overcame compared techniques. I succeeded also in designing of maze model, where agent was able to perform multistep tasks.
|
|
|
Analysis of the use of social sites as a key communication tool of the Czech Athletic Federation and their's change over time
Kolář, Adam ; Turková, Kateřina (advisor) ; Chladová, Marta (referee)
In my diploma thesis, I deal with social networks - specifically the social networks of the Czech Athletic Federation. It uses them as one of his main communication channels. Several quantitative and qualitative methods were used in the research. The main goal of the resarch was to compare social sites of the Czech Athletics Federation in 2019 and 2022. It was found that the main goal of the Czech Athletic Federation remains the same in the long term - to inform about the results of Czech athletes. Otherwise, however, social networks have undergone a transformation. The biggest difference over the years can be found in Instagram Stories, which the Federation has taken care of and significantly improved them over that three years. Media routines have also changed. Czech Athletics Federation tried to not be outdated, which is why a TikTok account was made and Instagram Reels were regularly created. The biggest failing can be seen in posts on Instagram, which practically copied Facebook. Overall, however, the year-on-year change on social networks can be seen positively, with which corresponds the growth of fans.
|
|
|
Hedging of a bond portfolio
Kolář, Adam ; Stádník, Bohumil (advisor) ; Kábrt, Tomáš (referee)
The aim of this bachelor thesis is to describe, apply and subsequently analyse the effectiveness of several hedging strategies used when cross-hedging a bond portfolio with futures contracts. The theoretical part of this work is devoted to a general description of characteristics of bonds and futures and, also to basic knowledge of interest rate risk hedging. In the empirical part of this thesis, a bond portfolio is complied, utilising market data, and subsequently it is secured with futures contracts using three different hedging strategies. At the end of the practical part, the effectiveness of each applied strategy is compared. The results of the empirical analysis support, that using the applied strategies when cross-hedging a bond portfolio with futures contracts is in most cases effective.
|
|
|
Travelling Salesman Problem
Kolář, Adam ; Rozman, Jaroslav (referee) ; Zbořil, František (advisor)
The aim of this bachelor's thesis is to design a testing environment for the traveling salesman problem and compare the effectiveness of different approaches to the solution. The first part discussed the possibility of genetic algorithms, depending on the setting of a crossover, mutations and population size. In the second part, there is the same problem using two types of neural networks. The representative of the self-learning net was chosen Kohonen neural network. Hopfield neural network represents a method of minimizing the energy function with fixed coefficients. At both neural networks, there were described possible advantages and disadvantages. In the end, all the findings were interpreted in a global context.
|
|
|
Coevolution of Cartesian Genetic Algorithms and Neural Networks
Kolář, Adam ; Král, Jiří (referee) ; Zbořil, František (advisor)
The aim of the thesis is to verify synergy of genetic programming and neural networks. Solution is provided by set of experiments with implemented library built upon benchmark tasks. I've done experiments with directly and also indirectly encoded neural netwrok. I focused on finding robust solutions and the best calculation of configurations, overfitting detection and advanced stimulations of solution with fitness function. Generally better solutions were found using lower values of parameters n_c and n_r. These solutions tended less to be overfitted. I was able to evolve neurocontroller eliminating oscilations in pole balancing problem. In cancer detection problem, precision of provided solution was over 98%, which overcame compared techniques. I succeeded also in designing of maze model, where agent was able to perform multistep tasks.
|
guest ::
RSS feed.