|
|
MLOSINT: Classifying Vehicle Losses in Ukraine
Kanát, Antonín ; Špelda, Petr (advisor) ; Střítecký, Vít (referee)
This thesis explores the potential of applying machine learning (ML) to assist with open source intelligence (OSINT) analysis. As the shared input of both disciplines, data is the primary lens through which the topic is examined. To understand the entire process of deploying an ML model from data collection to analysis, an image classifier of Russian vehicle losses in the invasion of Ukraine was trained and tested. Trained on a dataset of over 50,000 labelled images from the WarSpotting database, the classifier achieved a decent accuracy of 79% on evaluation data on the five most populous categories of images. On testing data from a later period, the performance dropped to 62%. One explanation offered is that the static frontlines and the prominence of drones led to most of the recent imagery being aerial, while the training data was captured mainly from the ground. That result demonstrated how inevitable changes, even in seemingly well-curated data, can lead to the low performance of ML models in deployment. Beyond changes on the battlefield, deeper data issues came to light, including the cascading effects of early data management decisions and dataset imbalance. Overall, current image classification methods do not work well on the noisy data available.
|
|
|
Analysis of the Depiction of Ukrainian Women in The First Months after the Russian Invasion of Ukraine
Vlčková, Tereza ; Pavlík, Petr (advisor) ; Gheorghiev, Olga (referee)
This diploma thesis regards the media construction of the portrayal of Ukrainian women in the context of the war in Ukraine. From the very beginning, the war was heavily covered by media, and gendered stereotypes began to manifest itself in the early days. In the theoretical part, therefore, the work deals with the construction of gender roles in war conflicts and sexualized violence, as well as the Russian invasion of Ukraine. The methodological part then presents research methods. These are quantitative content analysis, which aims to find out which topics are most often associated with Ukrainian women. The sources for quantitative analysis are articles from serious and tabloid online servers that contain selected keywords. From the sample are for critical discursive analysis selected articles with typical and also with deviating framing. The aim of the research is to map what framing of texts and setting of the agenda are the media using to portray the situation of Ukrainian women and what ideological structures this news framing has. Another research question is how the construction of masculinity and femininity in a war context is reflected in individual articles.
|
|
|
Impact of Market Uncertainty on Stock-Bond Return Relation
Rulíšek, Filip ; Čech, František (advisor) ; Teleu, Saida (referee)
In this thesis, we investigate the relationship between stock and bond returns in the US market from January 2018 to May 2023, with a specific focus on the impact of market uncertainty on this relation. Employing the rolling window cor- relation method, we examine the dynamic correlation between these two assets, using the S&P 500 Index and the US 10-Year Treasury Price Index. The results show, on average, a negative correlation on both monthly and quarterly basis. On a monthly basis we also observed highly fluctuating patterns. Additionally, the findings presented herein demonstrate that both the level and changes in stock market uncertainty, measured by the CBOE Volatility Index, negatively affect the relationship between stock and bond returns. On average, during times of increas- ing market uncertainty, investors tend to shift their funds from risky stocks toward safer bonds, while periods of low market uncertainty are usually characterized by the opposite trend. We carried out the same analysis for 11 stock market sec- tors separately. Interestingly, this analysis revealed that the relationship between the returns of these sectors and government bond returns varies. While the ma- jority of sectors exhibit the same negative correlation as the overall market, few sectors, such as Utilities and...
|
|
|
Ukraine War Through the Eyes of Czech Reporters
Levíčková, Žaneta ; Osvaldová, Barbora (advisor) ; Just, Petr (referee)
The bachelor's thesis deals with war reporting from the war in Ukraine and its specifics through the eyes of Czech reporters who were present in the country themselves. The objective is to look at the work of journalists in this armed conflict from their viewpoint because they are the most knowledgeable on the topic and they deal with war reporting long-term and comprehensively. Based on the data obtained, the work aims to clarify what obstacles the reporters encountered during reporting and what caused them. The research - semi-structured in-depth interviews - was analyzed using the grounded theory method. In the form of interviews, the insights of six reporters who reported on the war in the period of February- August 2022 are put into context. At the end of the work, the main obstacles that journalists encountered while reporting on this war are stated, as well as the proposal of possible solutions of how they can be eliminated in the future.
|
guest ::
