Contributions in Merlin

Federated learning has become increasingly more popular due to limitations of the traditional machine learning methods regarding the data privacy. In addition due to technological evolution, the data volume in general has increased by a lot. Mobile devices are capable of storing more and more data. While traditional machine learning methods struggle to deal with these concerns, federated learning emerged from these problems. Two main approaches have been mainly used namely Centralized and Decentralized Federated Learning. The former one has gotten much more attention in comparison with its counterpart and thus possesses many aggregation rules which are resistant to attacks. The goal of this thesis is to propose a new aggregation rule which is resistant to attacks against the machine learning model for the decentralized setting to fill a gap where the research has no reached yet. This is done by extending an existing framework fedstellar, for federated learning. The case studies as part of the evaluation evaluate the algorithm on performance and resource consumption related metrics. They indicate that the performance of the algorithm depends on the situation. They also show the limitation of the algorithm and possibilities of expanding the algorithm to other applications.

Background: Gamification and storytelling have been implemented in slow breathing training apps, a proven effective way to reduce stress and improve mental wellbeing, with the goal of keeping users interested and engaged. However, it is widely acknowledged that users’ cultural background and socioeconomic status (SES) affect their perceptions of apps and therefore intentions to use it on a regular basis. Objective: Based on previous research on user engagement in digital health interventions, cross-cultural human-computer interaction (HCI) and technology acceptance theories, this thesis investigates the intention to use three distinct versions of a slow breathing training app (Stressless©: control group, Breeze©: with gamification, Tragic Kingdom©: with gamification and storytelling) between German and American users. Methods: Adult US and German nationals were recruited from online crowdsourcing platform Prolific. They were randomly assigned to one of the three condition groups and watched a 1-minute introduction video about the corresponding breathing apps in their preferred language. Afterwards, they reported intention to use in a survey, alongside with other attributes related to technology acceptance that are not evaluated in this thesis. As the dependent variable, aggregated intention to use were summed up based on four survey questions. Independent variables were participants’ nationality and app versions coded as categorical variables. Results: A total of 325 participants completed the study (153 German participants and 144 from US Americans). The results show that while national culture does not play a significant role in intention to use, the effect of app version is strong. Although not statistically significant, SES differences in intention to use is observed with no moderating effect of national culture. Conclusion: Even though most hypothesis cannot be rejected; this thesis still provides meaningful guidelines for future studies on cross-cultural and other demographic comparisons on digital health gamification design.

Knowledge workers profit immensely from modern technologies and use them for many aspects of their work. But these technologies also come with drawbacks, as people are now more than ever expected to be available at all times via their work devices. This can lead to blurring boundaries between ones work and private life, which in turn can have negative effects on the ability to psychologically detach from work and enjoy non-work time. In this study we introduce the concept of artefact-based reflection, a method through which knowledge workers can reflect on so called work-artefacts, like tabs, files or e-mails, that they worked with throughout their day and "clean up their workplace" at the end of their workday. Based on existing research and this new concept we developed Downtime, an application which aims at helping knowledge workers in facilitating psychological detachment through artefact-based reflection. We then conducted a user study over two weeks to evaluate the effects of our application. Our findings suggest that knowledge workers can increase their detachment from work and create mental boundaries between their work and non-work lives through reflecting upon the work-artefacts they used that day, especially when they were not actively seeking detachment from their work beforehand. Further research is necessary to evaluate the long-term effects of artefact-based reflection on a broader range of knowledge workers.

Polkadot’s validator selection process employs an iterative algorithm, which is dependent on the size of the staking system. As Polkadot’s staking network is growing, I propose a machine learning alternative approach to the current implementation, that is more independent of scale. The algorithm, the sequential Phragmén, aims to reduce a graph of nominator-validator edges to a subset of validators, the active set, and distribute the stake backing them, as evenly as possible. The goal of this thesis is to produce superior results, consequently improving the overall security or to provide solutions of equal quality in faster time. In order to achieve the goal, a pipeline is setup, that gathers data and transforms it such that it is suitable for machine learning models. Predictions are made, which are adjusted to fit the requirements set by Polkadot. The adjusted results are scored and ultimately compared to the solutions discovered by sequential Phragmén. An analysis of the training data reveals, that the active set remains highly static, with only 10 validators on average changing from era to era. This lack of diversity raised concerns regarding potential attack vectors for adversaries. Furthermore, it was observed that many nominators are acting inefficiently. Many of them do not execute their right to nominate up to 16 validators, which would maximize their chance of having a validator included in the active set. Additionally, many of them include validators, which are not eligible targets. This occurs since nominators frequently ignore their duty to actively tend to their validator preferences. They set them once and do not update them. Eligible validators become inactive (intentionally or unintentionally) and consequently remain as part of the nominators preferences. The prediction task was split up into three models: The first model predicts the next active set, the second model predicts the sum of stake each validator receives and the third predicts the individual stake distribution. The results show, that the first two models are trained well and produce satisfactory results. However, the learning curves of the third model reveal a bias, which make the predictions suboptimal. The source of the bias is likely the substantial changes in target values introduced by a slight shift of active set. We conclude that it is unlikely to outperform the sequential Phragmén using a supervised approach under the described conditions. Therefore, we recommend exploring an unsupervised approach for further research. Furthermore, we recommend the development of a tool for nominators, that could increase the convenience and the security of the overall staking system as a consequence.

Incorporating Pedagogical Conversational Agents (PCAs) into learning environments has shown promising benefits for learning. However, most of the research has been performed in the field of academia. Building on a previous paper, this study focuses on the application of PCAs for organisational workplace learning. Having a strong Corporate Identity (CI) has become a competitive advantage for companies. In sectors, like public administrations that have a distinct service side, training staff on CI can be beneficial for the organisation. This thesis is part of a bigger research project that designed and implemented a CI Training for two public administrations in Southern Germany. Using Action Design Research, this thesis focused on designing a learning platform and learning videos that aimed to train staff on external communication, based on existing theories about multimedia learning and practice-inspired research. The main role of this thesis in the project was to design the usability of the learning platform, the experience of the learners, as well as the learning videos. So far, no uniform guidelines for the design of learning platforms that include a Pedagogical Agent (PA) could be found. The aim of this thesis was therefore to investigate the interaction of the agent, the platform, and the learner, thus contributing towards answering the question how a successful workplace training environment, that is supported by a PCA, can be designed. The results showed that incorporating a PCA into the learning environment increased the overall usability of the platform, especially the perceived control that the learners have over the interactions with the platform. The agent initially motivated the participants, however, this motivation declined as the agent did not perform to their expectations. No effect of the PCA-supported learning treatments on self-efficacy or self-determination could be found, in comparison to traditional e-learning without a PCA. Based on the results and on the qualitative feedback from the participants, design principles for PCA-supported learning environments were formulated. The data and results of this thesis can be used for further iterations of the overarching research project. Simultaneously, it can serve as a source of information for the design of other PCA-supported learning environments.

Detecting and solving sleep disorders can significantly impact society and the economy in general. The polysomnogram is the gold standard exam for diagnosing sleep disorders. Manually annotating the patient's sleep has limitations, including its time-consuming and tedious nature, lack of reliability, sensitivity to the setup of different clinics, and motion noise. This work tests the ability of neural network models to be faster and more reliable than manual scoring by incorporating temporal information in the training setting and changing the model architecture. The study concentrates on algorithms that are robust to the setup of different clinics and fair to diverse populations, using an intelligent combination of the most used datasets in experimental settings: the Sleep-EDF and the MASS datasets. We first analyze the ability of the automated classifier to handle data from different sleep centers and patient groups by experimentally testing loss functions and other crucial model parameters across datasets. Then, we incorporate temporal context in the data samples by concatenating previous sleep epochs to the current sample. We show that our model trained on longer temporal context performs equally to many of the analyzed manual sleep stage scoring conducted by expert technicians and is superior to some state-of-the-art models we analyzed.

This paper investigates the relationship between stock liquidity and the corresponding ADR shares discount using a sample of cross-listed firms in both the A shares market in China and the ADR market in the USA. The liquidity hypothesis is examined by introducing the market depth variables. The result of this study indicates that a smaller difference in stock liquidity between A shares and ADR is related to a lower corresponding discount in the ADR market. Such an effect is validated by the subperiod check. Overall, this study highlights that market liquidity of cross-listed stocks explains a proportion of the variation in price disparity that long existed in the ADR market. Meanwhile, a stock-specific factor for the price disparity is also proposed in this paper.

null

Media analysis reflects the most important facts about public opinion and the actors involved in the most current topics. It is beneficial for the government and the public administration to identify the information gaps in the public media to intervene and assist in specific areas concerning the reception and acclimatization process of Ukrainian refugees in Switzerland. This study aims to explore the potential of combining the three Natural Language Processing (NLP) techniques, Named Entity Recognition (NER), Sentiment Analysis (SA), and Topic Modeling (TM), with Machine Learning (ML) models to facilitate decision-making by the authorities. The data used for this study consist of German- and English-speaking articles retrieved from Swiss news media related to the Russian-Ukrainian war. The data includes articles between February 2022 and January 2023. The research approach is split into three main parts, data exploration, the modeling phase, and the evaluation of the models. During data exploration, the data was preprocessed and filtered in relation to Ukrainian refugees. In the modeling phase, the used ML models were introduced, fine-tuned, and applied to the data, leading to the final results. Using the three abovementioned NLP techniques, the most common topics, the article's sentiment over time, and the most common entities in the data could be identified. SA reveals a change of 5% from positive to negative articles regarding the total sum. TM presents the most common topics related to the Ukrainian refugee crisis in Switzerland. NER uncovers the most affected actors, locations, and organizations impacted by the crisis. In the evaluation phase, the model's performances were analyzed, which resulted in a remarkable accuracy score of 96.5% for the NER model, an average accuracy score of 62.5% for the SA model, and a coherence score of 0.65 for BERTopic, the TM model. To conclude, the research shows the potential of using ML-based NLP techniques on news media data to extract beneficial facts from a huge amount of data regarding the case of the Russian-Ukrainian war.

This paper examines the impact of the government’s partial guarantee loan provision (SBA 7(a) program) on small business lending during financial constraint periods. Our goal is to determine if access to loan securitization reduces banks’ incentives to screen borrowers, potentially leading to weakened screening standards. Our regression model and matching procedure support the idea that banks with high originate-to-distribute (OTD) ratios have sufficiently screened risky borrowers, which is supported by the evidence that high OTD banks have a wider loan pricing residual distribution. After looking into capital structure, we further confirm that banks are using SBA loans and the responsive secondary market to save regulatory capital, rather than diluting screening standards. This research provides insights into understanding government-guaranteed lending programs, the role of soft information in lending decisions, and the effects of loan sales on loan performance. Keywords: Originate-to-distribute, Small business lending, Financial crisis, Screening incentive, Government-guaranteed lending