STEM students in LMS-supported courses often face self-regulation challenges and avoid seeking help online despite high motivation. Existing learning analytics (LA) interventions frequently fail to address these issues, as they are not grounded in students’ behavioral and psychological barriers. This paper presents a sequential mixed-methods study (N=106, n=14) investigating self-regulation, motivation, and support behavior alongside academic stress areas (self-study, exam preparation, group work, instructor interaction) across four STEM courses. Findings reveal a disconnect between strategic knowledge and implementation, driven by procrastination and fear of appearing incompetent in public forums. Based on these results, we derive six design principles for LA-based digital nudging interventions, such as structured re-entry points after inactivity and just-in-time reminders keyed to temporal risk. Each principle is aligned with specific LMS log data and behavioral mechanisms. The resulting framework translates empirically identified student challenges into log-data-based design requirements for scalable and ethically grounded nudging interventions in LMS.

MOTIVATION: Single-cell transcriptomics sequencing is used to compare different biological processes. However, often, those processes are asymmetric which are difficult to integrate. Current approaches often rely on integrating samples from each condition before either cluster-based comparisons or analysis of an inferred shared trajectory. RESULTS: We present Trajectory Alignment of Gene Expression Dynamics (TrAGEDy), which allows the alignment of independent trajectories to avoid the need for error-prone integration steps. Across simulated datasets, TrAGEDy returns the correct underlying alignment of the datasets, outperforming current tools which fail to capture the complexity of asymmetric alignments. When applied to real datasets, TrAGEDy captures more biologically relevant genes and processes, which other differential expression methods fail to detect when looking at the developments of T cells and the bloodstream forms of Trypanosoma brucei when affected by genetic knockouts. AVAILABILITY AND IMPLEMENTATION: TrAGEDy is freely available at https://github.com/No2Ross/TrAGEDy, and implemented in R.

A powerful new complement to traditional synchronous teaching is emerging: intelligent tutoring systems. The narrative: A learner interacts with a digital agent. The agent reviews, selects and proposes individually tailored educational resources and processes – i.e. a meaningful succession of instructions, tests or groupwork. The aim is to make personal tutored learning the new norm in higher education – especially in groups with heterogeneous educational backgrounds. The challenge: Today, there are no suitable data that allow computer-agents to learn how to take reasonable decisions. Available educational resources cannot be addressed by a computer logic because up to now they have not been tagged with machine-readable information at all or these have not been provided uniformly. And what’s worse: there are no agreed conceptual and structured models of what we understand by “learning”, how this model-to-be could be implemented in a computer algorithm and what those explicit decisions are that a tutoring system could take. So, a prerequisite for any future digital agent is to have a structured, computer-accessible model of “knowledge”. This model is required to qualify and quantify individual learning, to allow the association of resources as learning objects and to provide a base to operationalize learning for AI-based agents. We will suggest a conceptual model of “knowledge” based on a variant of Bloom’s taxonomy, transfer this concept of cognitive learning objectives into an ontology and describe an implementation into a web-based database application. The approach has been employed to model the basics of abstract knowledge in engineering mechanics at university-level. This paper addresses interdisciplinary aspects ranging from a teaching methodology, the taxonomy of knowledge in cognitive science, over a database-application for ontologies to an implementation of this model in a Grails service. We aim to deliver this web-based ontology, its user-interfaces and APIs into a research network that qualifies AI-based agents for competence-based tutoring.

CNNs are characterized in particular by the ability to independently learn suitable features from a given data set. However, the resulting latent space is optimized for the given training data. Especially for tasks that require a high generalization ability, like e.g. the segmentation of single cells in a microscopic image across various experiments, these specific solutions might not offer optimal results. In this work, we improve generalization with an additional unsupervised training step that operates in the latent space. First experiments with the Kaggle cell segmentation competition data show a strong improvement in the generalization of acquired knowledge when using a soft- and hard-competitive Neural-Gas algorithm for deep clustering with a standard CNN architecture.

On the way to understanding complex perception tasks based on psychophysical data alone, we propose a general framework using multivariate analysis methods to derive a low-dimensional mapping of the underlying perception space. Psychophysical data can be interpreted in two fundamentally different ways: That is, the characterization of stimuli (e.g. colors) using several verbal descriptions (e.g. bright) - a stimuli-as-points view - and conversely, the characterization of the given verbal descriptions by several stimuli - a descriptors-as-points view. We argue that only the latter view enables us to reach an objective mapping of the perception space. For color perception, we show how it is possible to derive objective maps of the perception space, just based on non-comparative verbal descriptions of color stimuli. We also give an example where we analyze odor perception in the same way to derive a quantitative map of odor perception, a perceptual space that is still fairly unknown in its detailed structure.

BACKGROUND: Results of phylogenetic analysis are often visualized as phylogenetic trees. Such a tree can typically only include up to a few hundred sequences. When more than a few thousand sequences are to be included, analyzing the phylogenetic relationships among them becomes a challenging task. The recent frequent outbreaks of influenza A viruses have resulted in the rapid accumulation of corresponding genome sequences. Currently, there are more than 7500 influenza A virus genomes in the database. There are no efficient ways of representing this huge data set as a whole, thus preventing a further understanding of the diversity of the influenza A virus genome. RESULTS: Here we present a new algorithm, "PhyloMap", which combines ordination, vector quantization, and phylogenetic tree construction to give an elegant representation of a large sequence data set. The use of PhyloMap on influenza A virus genome sequences reveals the phylogenetic relationships of the internal genes that cannot be seen when only a subset of sequences are analyzed. CONCLUSIONS: The application of PhyloMap to influenza A virus genome data shows that it is a robust algorithm for analyzing large sequence data sets. It utilizes the entire data set, minimizes bias, and provides intuitive visualization. PhyloMap is implemented in JAVA, and the source code is freely available at http://www.biochem.uni-luebeck.de/public/software/phylomap.html.