Electricity is a quite unique commodity. Due to the economically non-storable nature of the commodity that electricity is, the constant balance between consumption and production, weather effects, such as temperature, wind speed, solar intensity etc, and the intensity of everyday and business activities, e.g. holi- days, weekends, on- and off-peak hours etc., the price dynamics of this commodity are quite unique as they are not observed in any other market. This extreme price volatility has forced market participants to hedge volumes as well as price risks. Naturally, electricity price forecast models are of great interest to portfolio managers.
Current data-driven models are combined with expertise of traders due to the considerable uncertainty of these models. We aim to subject trader expertise to a transparent methodology using structured expert judgement. During this study, we conduct two elicitations whose variables of interest concern average day- ahead spot prices for 2025, 2030 and 2035. We distinguish between baseload and peakload price. The first elicitation uses assessments regarding current Dutch day-ahead electricity spot prices to forecast forecast the variables of interest. We forecast the average day-ahead electricity spot price for 2025, 2030 and 2035. The second elicitation uses assessments regarding data on past, current and future developments in the Dutch electricity markets to forecast forecast the variables of interest. The second elicitation results in a decision maker with a higher price forecast than the decision maker of the first elicitation. Uncertainty regarding the forecast is higher for 2025 and 2030 during the second study. Uncertainty is equally high regarding the peakload price forecast for 2035 in both studies, however the range shifts.

In the process of drilling wells to produce hydrocarbons, an exploration strategy is used to determine which wells should be drilled and in which order. This strategy is vital, as a suboptimal drilling sequence will lead to more expenses and fewer gains.
Furthermore, the wells considered in most exploration strategies are geologically
dependent. Thus, a realistic model of these dependencies will be beneficial and
contribute to a more reliable optimal drilling strategy.
Previous research has shown that modelling similarities between the geological
properties of prospect wells in the same region and updating the drilling strategy dynamically after more information is available can add much value. However, the currently developed models are not realistic enough to use in practice. Previous models separated the process into two parts, first modelling the geological success with deterministic rewards and then applying dynamic programming to obtain the sequential drilling strategy.
This thesis proposes a new model with the addition of flexible uncertainty and
dependency models between the geological characteristics in the hydrocarbon volume calculation of a reservoir. For this purpose, non-parametric Bayesian networks in combination with copulae are used. This model and the input format of the required data align with how data is gathered in the oil and gas industry. The dependencies are modelled based on simple expert assessments.
After the new model for the geological properties of prospect wells and their dependencies is constructed, a decision policy that hinges on the observations of previously drilled wells is constructed. Previously used dynamic programming becomes infeasible with continuous variables and with a large number of wells. Therefore, the reinforcement learning algorithm Q-learning is used, which is better equipped to handle these more complex models.
The analysis shows that Q-learning achieves similar results to dynamic programming. Afterwards, the new model is applied to a case based on real oil and gas industry data. The results show that this case does not lend itself to an improvement when the drilling strategy is sequentially updated as the decisions compared to the original drilling strategy did not change. This occurs because the prospects all have high positive expected rewards and should be drilled, regardless of the other wells’ outcomes and dependencies.
Another case with synthetic data is explored with more favourable characteristics
that, in theory, should lead to the result of the policy from the new model and Qlearning outperforming the current industry policy. However, this was not shown yet within the time constraints of this thesis, as the Q-learner should be run for a more extended period and tuned according to the case.
While the model is created for dynamic sequential decision making in the application of drilling wells, this method can easily be adapted to fit any other application where decisions should be updated sequentially, and the
dependencies can be modelled through a Bayesian network.
In conclusion, the key contributions of this thesis are the creation of a new model for the dependencies between geological properties of a hydrocarbon reservoir and the use of the Q-learning reinforcement learning algorithm to compute the sequentially updated drilling strategy as an alternative to dynamic programming.

Applying multiple structure learning algorithms like hill-climbing, incremental association and grow-shrink, to investigate their robustness, predictive capabiliy and goodness of fit on multiple discrete Bayesian networks and Gaussian Bayesian networks.

Online customer reviews on products have become a large part of marketing intelligence in recent years. These documents are a source of information on what aspects of a discussed product can be improved upon. These aspects are named drivers. CQM, the company in which the internship for which this thesis was written took place, has developed a tool in which reviews are manually annotated for a fixed set of drivers. The result is that for each driver, each review can be assigned a driver score. The driver score can be a positive or negative number, indicating with what sentiment the reviews discusses the driver, or left blank, meaning that the review does not discuss the driver. The goal of this work is to (partially) automate this process, so that, given the review, the driver scores can be predicted. The first step towards achieving this is creating a binary classification problem for every driver, where a binary variable can, for example, indicate whether a review discusses the driver or not. A step further is multinomial classification, where one can also distinguish whether a driver is discussed with positive or negative sentiment. The reviews are represented as variables, with every variable representing the use of a word. In this thesis, different forms for these variables are experimented with. With every variable representing a unique word, a large number of distinct predictors is offered for the classification problem. Given these predictors, two types of models are considered to solve the classification problem: Elastic net models and random forests. Both types of models need to be adapted for the class imbalance before they can be used for the classification problem of predicting the driver scores. The results of these models are evaluated using the area under the curve (AUC) and for the multinomial problem a multinomial generalisation, the AUCμ. These measures are chosen, because they are effective at evaluating the performance of our models in the context of the class imbalance. The results were ultimately evaluated for various variable forms and models. For the model we found that a random forest adapted to use stratified bootstrap samples to grow decision trees gave strong performance, especially when combined with variables that were given an indicator function form or normalized tf-idf form.

The objective of this research is to understand what smart charging can bring business and society at large. In close collaboration with the largest fleet operator in the world and a commercial aggregator, the impact of smart charging on both cost reduction and carbon reduction was simulated for an electric vehicle (EV) fleet in 2018. The simulation is designed to quantify the cost reduction of EV smart charging in The Netherlands as realistic as possible. Besides cost reduction quantification, the objective is to create a better understanding of what variables influence smart charging cost reduction. This is done via a statistical analysis of the smart charging simulations output. Furthermore, this research also has the objective find the direct impact of smart charging on carbon intensity of the electricity used for personal transport.
In this research, an EV aggregators perspective is leading. The EV aggregators could utilizing available flexibility in an EV fleet to deliver flexibility services. The strategy chosen to simulate is based on day ahead market optimization and passive balancing on the imbalance market. The EV fleet is assumed to be an isolated portfolio handled by the balance responsible party (BRP). The average synthetic load profile over 2018 was €41,56 per MWh and this is used as benchmark to quantify the smart charging savings in the simulations.
Different smart charging simulation set-up scenarios are designed and executed. In all simulations a real-world charging data set with 300.000 historic charging sessions was used. For each session, a new smart charging profile is determined by the optimization algorithms. The session price and session carbon intensity is calculated for both the smart charging scenarios as the business-as-usual scenarios. To compare the results of the different smart charging set-up scenarios, the average session price of all session in that simulations is used. At the same time, the smart charging savings is calculated based on the defined benchmark. The findings within this thesis support the conclusion that the used smart charging algorithms work properly and could decrease the electricity purchase price in The Netherlands. Additionally we found that the carbon intensity of the charged electricity during the smart charging schedule decreases compared to a business as usual scenario. This is a direct result of a correlation between the carbon intensity in the grid and day ahead prices in The Netherlands. EV aggregators are able to add flexibility to the demand side of the electricity system by means of smart charging, if a strong price incentive is provided. If stakeholders across the mobility and the energy sector work together, a real-world commercial implementation based on the price incentives on day ahead market and imbalance market in The Netherlands is possible. In the statistical analysis, multiple regression models show a linear relation between three independent variables (the session duration, session volume and maximum power of the charge point) and two dependent variables (the average session purchase price and savings per session). The key insights from the models empowered three main recommendations to EV aggregators to optimize the smart charging savings in the future: 1) Encourage longer session length. 2) Encourage regular overnight charging sessions behaviour, independent from the charging needs. 3) Stimulate access to high charging power. The data showed compelling differences between the 20% BEVs and the 80% PHEVs and their results were separated accordingly in this research. In all simulation set-up scenarios are the PHEVs outperforming the BEVs in terms of a lower average session price and higher cost reduction. If the smart charging strategy is executed as proposed in this thesis, the EV aggregator is exposed to the day ahead market and imbalance settlements for its portfolio. The EV aggregator is able to decrease the electricity purchase price, while acting as BRP. The exposure to the markets brings significant risk. Collaboration with an electricity supplier or BRP could potentially increase the smart charging savings for the EV aggregator. Furthermore, other revenue streams to utilize flexibility could be investigated. If stacking different flexibility strategies is possible, it could increase the smart charging value in the future.

In this thesis we try to capture the dependence structure of the publications of a scholar and the citations of those publications via copulas. To do so, we will use a sample of Quebec re- searchers for who their publication amount as well as their citation amounts are known. We are provided with multiple variables concerning citation. We study the dependence structure be- tween these variables, with the aim of fitting copulas to this structure, by calculating correlation scores and visualising the structure. Copulas are functions that ”join together” one-dimensional distribution functions with a dependence structure, in order to represent joint distributions. The correlation scores are calculated across various ranges of the variables to provide us with a deeper understanding of the dependence structure between the variables.
Using Sklar’s theorem and some helpful functions in various packages in the software program R, parametric copulas fit the dependence structures of the various pairs of variables. Based on a Goodness-of-fit test, certain parametric copula models are rejected at a 5% significance level. Unsurprisingly, there are also dependence structures that can be well captured with a parametric copula.
Parametric copula families are not only used for fitting the data, but also for prediction. Since a good fitting model does not necessarily imply a good predictive model, we have also performed a validation analysis. The parametric copula models that are not rejected by the test at a 5% significance level are validated via k-fold cross validation. Part of the data have been used to fit the model and the remaining has been validated using a k-fold cross validation. It turns out that the best fitting copula model does not always perform well in term of prediction. That is, these copulas do not always perform best during the cross-validation.