|
|
Reinforcement Learning for Bomberman Type Game
Adamčiak, Jakub ; Beran, Vítězslav (referee) ; Hradiš, Michal (advisor)
This bachelor's thesis aims to develop, implement and train reinforcement learning models for a Bomberman-type game. It is based on Bomberland environment from CoderOne. This environment was created for education and research in the field of artificial intelligence. In this thesis I tackle the settings and problems of implementing agent into the environment. I used 2 policies (MLP and CNN), 2 algorithms (PPO and A2C) and 5 setups of neural networks for feature extraction with the use of libraries stable baselines 3 and pytorch. Total training time resulted in 1207 real-world hours, 4168 computing hours and 271 milions of time steps. Although the training was not successful, this thesis shows the process of implementing a reinforcement learning model into a Gym environment.
|
|
|
Media Framing: Transformation of Nursultan Nazarbayev's Image in the US Media
Tokayeva, Assem ; Miessler, Jan (advisor) ; Jeřábek, Hynek (referee)
The thesis deals with U.S. media coverage of Kazakhstan's first president Nursultan Nazarbayev between 2011 and 2022. Using Bourdieu's theory of journalistic field and Entman's concept of media framing, the content analysis of five different media outlets shows that while commercial newspapers (NYT, WP, WSJ) were more critical than non- profit online media outlets (Eurasianet, RFE/RL) regarding Nazarbayev and his regime during events that challenged his power; their overall coverage of the country, its leader and his legacy has been restrained or even credited Nazarbayev for various achievements. Differences in the degree of critical stance were also identified between the two non- profit media. Providing an overview of Kazakh government-funded lobbying information campaigns abroad, the thesis confirms the importance of research into the use of international media platforms by authoritarian regimes aiming at creating a favorable image abroad. Keywords: Nursultan Nazarbayev, Kazakhstan, media framing, lobbyism, U.S. media, RFE/RL, Eurasianet, NYT, WSJ, WP Range of thesis: 52 pages, i.e. 119,088 characters
|
|
|
Donald Trump's Attempts to Influence the Operation of Radio Free Europe/Radio Liberty and the U.S. Agency for Global Media
Mužíková, Natálie ; Sehnálková, Jana (advisor) ; Raška, Francis (referee)
This bachelor's thesis deals with the influence of Trump's policy on American broadcasting abroad. The aim of the work was an objective evaluation of the influence and role of Trump's policy on the Independent Global Media Agency (USAGM) and Radio Free Europe/Radio Liberty (RFE/RL). The bachelor's thesis briefly describes the history and development of RFE/RL and USAGM, as well as the relationship of Donald Trump with the media and especially the character of Stephen Bannon. The following part of the work introduces the character of Michael Pack, along with the controversies caused by his nominations and especially his subsequent work as the director of the USAGM. The result of the research is that even though the changes made by USAGM CEO Michal Pack were fundamental and in the short term paralyzed the agency, they were mostly perverted by the advent of the new administration. In conclusion, the idea of influencing the media, which is now very current, and the question of the future of the agency and broadcasting stations themselves are emphasized. The bachelor's thesis uses the procedures of qualitative research, analysis of existing literature and synthesis.
|
|
|
Reinforcement Learning for Bomberman Type Game
Adamčiak, Jakub ; Beran, Vítězslav (referee) ; Hradiš, Michal (advisor)
This bachelor's thesis aims to develop, implement and train reinforcement learning models for a Bomberman-type game. It is based on Bomberland environment from CoderOne. This environment was created for education and research in the field of artificial intelligence. In this thesis I tackle the settings and problems of implementing agent into the environment. I used 2 policies (MLP and CNN), 2 algorithms (PPO and A2C) and 5 setups of neural networks for feature extraction with the use of libraries stable baselines 3 and pytorch. Total training time resulted in 1207 real-world hours, 4168 computing hours and 271 milions of time steps. Although the training was not successful, this thesis shows the process of implementing a reinforcement learning model into a Gym environment.
|
guest ::
