Working with pneumatics can be done using a Programmable Logic Controller (PLC), which commands the individual actuators. To start, the movements of each component are described in a displacement diagram. From this diagram, ladder logic can be extracted, which is a set of conditions with corresponding actions. These actions are executed when the conditions hold. This ladder logic is the programming language of PLCs and makes it possible to control pneumatics. De Techniekschool provides courses on these techniques and pneumatics. This report describes the research, design and implementation of a software application to aid students in their education of pneumatics. The main goal is to allow users to create a displacement diagram and then generate the corresponding ladder logic based on this diagram. This will allow users to understand how this conversion works and help them in understanding PLC programming. In the research phase, various designs were evaluated. The implementation of the compiler was thought out and possible designs for the Graphical User Interface (GUI) for creating the diagram were explored. In the case of the GUI, the final implementation differs from the initial chosen design. This is because, during the project, it was discovered that the design did not work as well as intended, thus the switch was made. To guarantee the correct behaviour of the product, each individual component was tested on its own with unit tests. Besides a compiler and the ability to draw displacement diagrams, features of Stepladder include the option to save and open work, and the option to undo and redo actions. It is also possible to show the signals which the PLC will receive and send. These signals already play a vital role in the conversion from displacement diagram to ladder logic and the generated signals can help to understand this conversion. Desirable features which are not yet implemented are the support for multiple languages and improvements to the accessibility of the application. Also, a more advanced compiler is desired, because the current compiler does create ladder logic according to a valid technique, but it is not the optimal solution. For learning purposes, an optimal conversion may be desired. Proper user testing is also desired to verify the design and whether the application is easy to use. To conclude, the final product allows the user to create displacement diagrams and generate ladder logic. This can be used to study the conversion needed in PLC programming. The final product meets the design goals set in the design process. Besides the design goals, the product also satisfies the initial requirements. There are some features the client could implement but these are not required.

Synthetic lethality (SL) is a relationship between two genes, exploited for targeted anti-cancer therapy, whereby functional loss of both genes induces cell death, but the functional loss of either gene alone is non-lethal. Computational prediction of SL gene pairs is sought after because it is expensive to do lab screening for SL. Existing SL labeled pairs from wet- lab experiments often focus on specific genes or pathways, resulting in notable selection bias. Current SL prediction methods ignore this bias when training on available SL labels, and fail to generalize if test sets follow a different selection bias. One way to mitigate bias is to incorporate unlabeled pairs during model learning. However, conventional semi-supervised methods such as self-training can reinforce bias by adding confidently pseudolabeled pairs, which tend to be most similar to previously included samples. We present DBST, a self-training strategy that addresses the issue by promoting diversity in the selection of pseudolabeled samples. This is achieved using metric learning to find a class-contrastive representation of the feature space, based on which DBST selects diverse (or dissimilar) pseudolabeled pairs. In results for five cancer types, semi-supervised models, including DBST, delivered improved SL prediction performance over the supervised model. Additionally, DBST successfully incorporated unlabeled samples that were more dissimilar among them compared to standard self-training. In experiments with differing biases between train and test sets, DBST showed a slight improvement in performance compared to the supervised model.