Negotiation is a challenging process for people, which often results in suboptimal agreements between the negotiating parties. This issue leads to lost benefits that one of the negotiating parties could have obtained. To counteract this drawback artificially intelligent negotiation agents are developed. Their goal is to help negotiating parties increase the overall benefits from a negotiation. However, the main drawback of such agents is that people need to trust the agents in order to use them. In this research paper we investigate whether a negotiating agent with a hardliner negotiating style would be trusted more than a negotiating agent with a conceder negotiation style. We answer this question by conducting an user study, where we measure participants' trust in the two negotiation agents and compare the results. From the results we concluded that the assumption is correct for the sample group that participated in the user study.

This research experiment aimed to investigate the level of trust placed in an AI negotiation assistant paired with a truthful explanation of their negotiation strategy versus an opposite explanation within the Pocket Negotiator platform. A between-user study involving 30 participants was conducted to assess participants’ trust perceptions based on the presentation of different explanations about the negotiation assistant. After receiving an explanation
of the negotiation strategy used by the assistant, participants went through a bilateral negotiation on the Pocket Negotiator platform against a robot after
which they completed a questionnaire to evaluate their trust in the assistant. The results were insignificant (p > 0.05) and therefore no conclusion could be drawn about the difference in the participants’ trust in the assistant with a truthful explanation and the assistant with an opposite explanation.

With the prevalence of artificial intelligence recently, more attention is being drawn towards the collaboration between humans and agents. Across the many fields where such an agent can be employed, we are going to specifically examine the domain of negotiation. A critical part to ensure success in this human-agent collaboration is to establish trust between the user and the agent. When considering negotiations, different strategies can be employed by the negotiation agents, the specific negotiation agent we shall use in our study employs the so called "hardliner negotiation style", which can be thought of as taking an uncompromising stance. When introducing the agent, we will either show a neutral introduction which does not mention the choice of strategy or show the hardliner introduction which does cover the chosen strategy. We are going to investigate how the choice of introduction affects the trust of humans in the negotiation agent. Leading us to the research question: What's the effect of an agent using a truthful explanation of their hardliner negotiation style versus a neutral explanation on how much humans trust and rely on the agent to negotiate for them? A between-subjects study took place to answer the question which has indicated that there is no significant difference (U = 111, p = 0.97 two-tailed, α = 0.05) between the two different introductions.

Trust in negotiation agents plays a crucial role in their adoption and utilization. However, there is not enough research on what factors influence it. This paper aims to investigate how different explanations of a negotiation agent’s strategy affect human trust and decision-making. Specifically, it compares the effects of a truthful explanation and a neutral explanation on trust in a conceder negotiation agent. An experiment was conducted involving human participants interacting with a negotiation agent in an imaginary negotiation scenario. The participants were divided into two groups, with one group receiving a truthful explanation of the agent’s strategy, and the other receiving a neutral explanation. Trust levels were assessed through a questionnaire and the results were analyzed using an independent samples t test. No significant difference was found between the two groups, but further research needs to be done before conclusions are drawn.

Negotiations have an essential role in our lives as they help us to find mutually beneficial solutions and resolve conflicts. It leads to effective communication and collaboration between the involved parties. Negotiation among parties has high importance to have an outcome that is suitable for all. In such scenarios, negotiation agents can be used by the involved parties. Negotiation agents have explanations that indicate the behavior that the agent is going to follow in the process. This research paper focuses on trust between humans and agents regarding different explanations of the agents. In the research, the negotiation strategy conceder is used in the negotiation with an opposite and truthful explanation of itself. The results from the experiments are compared with each other using the scores of the participants which are measured through a questionnaire. By investigating the effect of different explanations, the research paper aims to answer the question ”What’s the effect of an agent using a truthful explanation of their conceder negotiation style versus an opposite explanation on how much humans trust the agent to negotiate for them?”. The goal is to achieve a trust connection between humans and agents. The main conclusion and the hypothesis of the research are ”people tend to trust less on an agent that does not do what it states in its explanation.” The results of the experiment, discussion of the results, and the conclusion can be found in sections 4, 5, and 7 respectively

An empirical study is performed exploring the sensitivity to hidden confounders of GANITE, a method for Individualized Treatment Effect (ITE) estimation. Most real world datasets do not measure all confounders and thus it is important to know how crucial this is in order to obtain comparable predictions. This is explored through the removal of confounders with varying strengths and by removing subsets of the confounders simultaneously. The sensitivity is measured through the change in Precision in Estimating Heterogeneous Effects (PEHE) and through the divergence in the estimation of Average Treatment Effect (ATE) from the GT. Experiments are performed on synthetic and semi-synthetic data. The number of removed hidden confounders increases the error and variability of predictions, both for ITE and ATE. The strength of the removed confounders does not show a conclusive relationship on the error metrics. The effect of removing confounders with different causal graphs is explored but fails to show any clear patterns due to the high variance of the results.

Distributed systems are networks of nodes depending on each other. However, each network can have multiple faulty nodes, which are either malfunctioning or malicious. Bracha's algorithm allows correct nodes inside the network to agree on certain information, while tolerating a certain amount of faulty nodes. Nodes exchange Send, Echo and Ready message types to inform each other of receiving the message and agreeing on its trustworthiness. The focus of this paper is to present the functional differences of exchanging message paths that messages have crossed instead of the message types. In conclusion, the BMP algorithm showed potential to outperform original Bracha's algorithm in two general cases: networks that have a low probability of successfully transmitting a message, and networks where nodes have a system of trust established which allows them to determine trustworthiness of other nodes. Otherwise, in general usage, original Bracha's algorithms appears to be superior in comparison to the BMP algorithm.

In this paper we will consider the Byzantine Reliable Broadcast problem on partially connected net- works. We introduce an routing algorithm for networks with a known topology. It will show that when this is combined with cryptographic signatures, we can use the routing algorithm to create an optimal amount of messages. We will introduce a few algorithms to estimate the routes required for this algorithm. Our simulations show that the amount of messages required for Byzantine Reliable Broadcast can be massively decreased by using this. We will also show that the method of choosing these paths has a huge impact, but come with their own trade- offs.

Causal machine learning deals with the inference of causal relationships between variables in observational datasets.
For certain datasets, it is correct to assume a causal graph where information about unobserved confounders can only be obtained through noisy proxies, and CEVAE aims to address this case.
The number of dimensions of the latent space modelled by CEVAE must be specified ahead of time, and this paper investigates the effect of this dimensionality misspecification on the performance of CEVAE.
Results support the idea that underspecification and overspecification both degrade the performance of CEVAE, but indicate that underspecification is worse, at least for the case with few confounders.
In general, the model does not always achieve best performance when the model dimensionality corresponds to the data dimensionality.
Finally, conclusions made on data with linear-Gaussian proxies are the same as those obtained with nonlinear-Gaussian proxies, which indicates these conclusions generalize over different datasets to some extent.

The large amounts of observational data available nowadays have sparked considerable interest in learning causal relations from such data using machine learning methods. One recent method for doing this, which provided promising results, is the DragonNet (Shi et al., 2019), which utilises neural networks in order to estimate average treatment effects in populations. The performance of the model, however, was not tested on datasets which contain low amounts of overlap between the treated and non-treated subpopulations, which makes it harder to accurately estimate treatment effects. Therefore, the goal of this paper is to investigate the performance of the DragonNet when used on datasets with (near) overlap violations. This has been done by looking at the mean absolute errors and variances of the estimated treatment effects and comparing these to other models. The results showed that the performance of the DragonNet becomes significantly worse compared to other models when large portions of the population suffer from low overlap. Additionally, the variance of the results also increases in these cases, making the results less reliable. From the obtained results, it can be concluded that it is best to choose another model for treatment effect estimation if relatively large amounts of overlap violations are suspected.

The purpose of this research is to analyze the performance of Propensity Score Matching, a causal inference method for causal effect estimation. More specifically, investigate how Propensity Score Matching reacts to breaking the unconfoundedness assumption, one of its core conceptual pillars. This has been achieved by running PSM on synthetic data that upholds the unconfoundedness condition, and then comparing these results with measurements obtained from running the algorithm on data with confounding features with varying contribution to other variable values and hiding these features individually or in progressively higher numbers. These results are also then compared to Linear Regression, a generic machine learning algorithm, for the sake of comparison of performance. The results obtained point to the observation that when hiding variables that only contribute to the main effect, treatment effect or treatment propensity calculation respectively, PSM performs with the same error no matter which of the three effects the hidden feature affects, making them equivalent in their error contribution. Additionally, it has also become apparent that in all experimental scenarios used in this work, PSM performed very similarly to Linear Regression and did not seem to offer any advantages over the latter in these specific situations.

Increasing digitalisation of society due to technical advancement has increased the appearance and size of cyber- physical systems. These systems require real-time reliable control, which comes with its challenges. These systems need reliable communication despite the presence of attacks or faulty processes and bad connections, which can disrupt the timeliness and correctness of these applications. Furthermore the scale of these systems are becoming increasingly larger as such the cost of connecting all processes that make up these systems become very high. This paper addresses the issues of reliable communication in these conditions while allowing for networks to be partially connected. We show that the protocol discussed can meet real- time requirements of these systems even under conditions of bad connections and low network connectivity.

Causal machine learning is a relatively new field which tries to find a causal relation between the treatment and the outcome, rather than a correlation between the features and the outcome. To achieve this, many different models were proposed, one of which is the causal forest. Causal forest is made up of a random forest, with a different estimation function in the leaf node, which means it suffers from the same problems, like being easy to overfit. The reason why honesty was introduced was to ensure mathematically that forests do not overfit as easily. This research however, only provided preliminary results and no real testing was done in terms of causal inference. In this paper three scenarios are tested where a comparison is made between a causal forest with and without honesty. Based on the results it seems that honesty does indeed help for trees to not overfit. However in a general setting it hurts the model as it only trains with half of the available data. This makes honest causal forest less accurate in general settings where there is not a lot of training data. In a setting where a large amount of data is provided it seems that honesty does not change the performance, meaning it creates a theoretical guarantee against overfitting with no repercussions for the performance.

Discovering the topology in an unknown network is a fundamental problem for the distributed systems that faces several backlashes due to the proneness of such systems to Byzantine (i.e. arbitrary or malicious) failures. During the past decades, several protocols were developed to allow a network to be resilient up to a certain number of Byzantine processes and guarantee a reliable communication between the correct processes. However, they still suffer from high complexity or unrealistic assumptions which makes them impractical. In this paper modifications made to one of the current state-of-art protocols are presented, modifications that can potentially reduce the load of the network and improve the Reliable Communication layer. We employ empirical analysis in order to prove that reliable communication can still be achieved with these modifications in place and to measure the performance. The findings of this study show that our improved protocol has a reduction of 18.38% in the number of exchanged messages.

Large textures that can provide realistic details are widely used in modeling, gaming, art design, etc. Texture synthesis is a way to create large textures based on a small sample pattern, which can be obtained by image examples or hand­drawn work by an artist. Different methods that aim to achieve better visual effects on reducing or avoiding seams and distortions during synthesis are proposed. It has been a popular topic both in computer graphics and computer vision for many years. However, most of the synthesis methods are automatic and take images as the input. There are few methods designed for artists who want to control and create patterns manually. The goal of our Diffusion Mosaics is to introduce a graphics tool that allows the artist to create tiles that can be seam­ lessly concatenated. In this way, non­periodic textures of arbitrary size can be produced at very low memory costs. We rely on a special kind of tiles called Wang tiles, which are squares with colored edges. Neighboring tiles should share the same edge color at the common border. The texture tiles are designed in such a way that if this constraint is fulfilled, the transition from one tile to the next will be seamless. It becomes possible to create a large non­periodic final texture. In our system, each tile can be filled with hand­drawn patterns using diffusion curves, which is a vector graphics primitive created by diffusing the given colors of defined Bezier curves, which has been proven to be a useful alternative to purely pixel­based design that artists typically rely on. They also match our application well, as, by adding border­color constraints to the tile, solving the diffusion process results in tiles that match the neighboring tiles naturally, hence avoiding any seams. The tool gives creative freedom to the artist that they can even draw the pattern outside the tile area, resulting in an automatic update of all other tiles when needed to ensure consistency. Finally, for more complex illustrations and back­ wards compatibility to software such as Photoshop, pixel­based images are also supported and can be integrated.

The AI World Cup is a virtual competition in which teams of five players compete in a football match. The defensive strategies for the goalkeeper in this environment are yet to be researched, however. In previous editions of the competition the participating teams use a basic goalkeeper that can only dive but not position itself. This project provides research into the use of different positioning techniques to improve the performance of the goalkeeper.
In the project two different positioning techniques are presented and evaluated. In the first technique, the goalkeeper moves on an imaginary arc in front of the goal and in the second technique the goalkeeper moves on a straight line in front of the goal. Moreover, teamworking techniques between a defender and the goalkeeper are used to create a better coverage of the goal. The evaluations for the positioning techniques consist of 250 shots shot by an attacker from random distances and angles. The evaluations indicate that both goalkeepers have an improvement in performance over the basic goalkeeper, used by the other teams. The goalkeeper positioning on a line has the best performance, with an increase in performance of 27\% over the basic goalkeeper. These results demonstrate the value of positioning techniques for a goalkeeper in the AI World Cup.

To push the boundaries of technology, the world cup football for robots, RoboCup, is organized on a yearly basis since 1997. To push the boundaries of artificial intelligence, a simulated version of the RoboCup, AI World Cup Football, is arranged yearly from 2017. This requires skillful attackers, defenders and goalkeeper. A large part of having a competent goalkeeper, is having a well-anticipating goalkeeper. This research will improve the goalkeeper’s level of anticipation by finding its own weak spots. By training an attacker against this goalkeeper, the weak spots can be determined. Based on this data the goalkeeper can estimate in which region of the goal a real opponent is likely to score and decrease that scoring chance by making a move. Heuristics and a deep neural network are used to estimate how likely the attacker is to make a goal. The results show that the attacker that uses the deep neural network has a chance of scoring a goal which is about 10 to 15 percentage points higher than a random shooting attacker. The deep neural network attacker is however outperformed by the heuristic based player by about 15 to 20 percentage points. Taking that into account, embedding the heuristic based attacker into the goalkeeper gives the goalkeeper a good sense of what areas of the goal it should cover to decrease the likelihood of conceding a goal.

Ambient occlusion is a popular rendering technique that creates a greater sense of depth and realism, by darkening places in the scene that are less exposed to ambient light (e.g., corners and creases). Ambient occlusion measures how geometrically occluded each point in the scene is and modulates the ambient light accordingly. In real-time applications, screen-space ambient occlusion approximations are used, due to their great performance and visual quality. However, these screen-space approximations only take the geometry currently visible on screen into account. This results in underestimated or missing ambient occlusion when geometry is hidden from view (e.g., hidden behind other geometry). Our proposal, stochastic-depth ambient occlusion, improves upon traditional screen-space ambient occlusion techniques by including information about geometry at multiple depth layers, using a stochastic transparency based approach. This allows us to efficiently gather the missing information in order to improve upon the accuracy and spatial stability of conventional screen-space approximations, while still maintaining the real-time performance. Our approach integrates well into existing rendering pipelines and we show how it can be generalized to work with different screen-space ambient occlusion techniques and extensions (such as bent normals and cones). We also demonstrate how multi-view stochastic-depth ambient occlusion can greatly improve upon the robustness of traditional multi-view ambient occlusion techniques. Furthermore, we provide an extensive analysis of the visual quality, robustness and performance of stochastic-depth ambient occlusion compared to conventional screen-space approaches.

The recommendation of tasks for newcomers within a software project throughgood first issues is being done within the domain of software development, such as onGithubplatform. These issues aim to help newcomers identify tasks that are suitablefor them and their level of expertise within the project. This thesis report investigatesthe effectiveness regarding developer onboarding and task completion ofgood firstissues by data mining a set of 105 repositories and manually analyzing at most 30good first issues and 30 initial commits per sampled project. It was found that, althoughgood first issues are effective at developer onboarding, and developers perceivegoodfirst issues as being useful, changes can be made to the types of tasks suggested asgood first issues to match the types of initial contributions made by newcomers. It wasalso found that developers with less than a year of experience favoreddocumentation-related contributions for their first commit to a project.