Claudia Plant

Understanding causal relationships in multivariate time series is essential across many scientific domains, especially in climatology and meteorology where complex dependencies drive extreme events. Existing tools often lack intuitive visualization, particularly for heterogeneous Granger causality applied to non-Gaussian data such as time series following exponential distributions. There is a need for an accessible, interpretable tool that helps scientists explore temporal dependencies and uncover causal structure in such settings. We present HMMLVis, an original visualization tool designed for multivariate and heterogeneous Granger causal inference (Heterogeneous Granger causality by Minimum Message Length). The tool supports causal discovery in time series with exponential distributions and offers an intuitive interface for exploring temporal relationships. While HMMLVis is applicable across disciplines dealing with time-series data, this work focuses on its use in climatological and meteorological contexts. We demonstrate HMMLVis on several applications involving meteorological phenomena affecting the upper and lower distributional tails, using datasets from renewable energy (wind and PV), air pollution, and the European Meteorological Network (EUMETNET) post-processing benchmark (EUPPBench) across different temporal horizons. Our results show that HMMLVis successfully recovers known causal relationships and additionally reveals previously unobserved temporal dependencies relevant to the specific cases examined. As an interpretable and user-friendly visualization tool, HMMLVis has strong potential to support climatologists, meteorologists, and other researchers working with complex time-series data. By enabling clearer insight into causal interactions, it contributes to more informed scientific understanding and facilitates knowledge discovery across multiple environmental and atmospheric science applications.

How to find a natural grouping of a large real data set? Clustering requires a balance between abstraction and representation. To identify clusters, we need to abstract from superfluous details of individual objects. But we also need a rich representation that emphasizes the key features shared by groups of objects that distinguish them from other groups of objects. Each clustering algorithm implements a different trade-off between abstraction and representation. Classical K-means implements a high level of abstraction - details are simply averaged out - combined with a very simple representation - all clusters are Gaussians in the original data space. We will see how approaches to subspace and deep clustering support high-dimensional and complex data by allowing richer representations. However, with increasing representational expressiveness comes the need to explicitly enforce abstraction in the objective function to ensure that the resulting method performs clustering and not just representation learning. We will see how current deep clustering methods define and enforce abstraction through centroid-based and density-based clustering losses. Balancing the conflicting goals of abstraction and representation is challenging. Ideas from subspace clustering help by learning one latent space for the information that is relevant to clustering and another latent space to capture all other information in the data. The tutorial ends with an outlook on future research in clustering. Future methods will more adaptively balance abstraction and representation to improve performance, energy efficiency and interpretability. By automatically finding the sweet spot between abstraction and representation, the human brain is very good at clustering and other related tasks such as single-shot learning. So, there is still much room for improvement.

In this work, we present HMMLVis, an original visualization tool for multivariate Granger causal inference. More precisely, for heterogeneous Granger causality to infer causal relationships in time-series following an exponential distribution. HMMLVis is easy to use and can be applied in any scientific discipline exploring time series and their relationships. In this paper, we focus on climatological and meteorological applications. The visualization tool is demonstrated on different types of applications related to meteorological events on the upper/lower tails of the respective distributions using a renewable energy (wind, PV), air pollution, and the EUMETNET postprocessing benchmark data set (EUPPBench) and different temporal horizons. We demonstrate that the HMMLVis method and visualization depicts the known causal and detects causal relations in the temporal dependencies which are additional important information for the respective cases. We believe that HMMVis as an interpretable visualization tool will serve climatologists or meteorologists and in this way it will contribute to knowledge discovery in these scientific fields.

It is generally accepted that collaborative information is important for the performance of recommender systems. It is also generally accepted that if this information is sparser, it impacts recommendation systems negatively. Various approaches have tried to lift this problem by employing side information. However, global patterns that can be provided by clusters of similar items and users or even additional information such as text are often not used together with collaborative information. We study the impact of integrating clustering embeddings, review embeddings, and their combinations with embeddings obtained by a recommender system. We study the performance of this approach across various state-of-the-art recommender system algorithms including graph-based methods. We highlight that graph structures are important with sparser datasets and both, in knowledge graphs with side information as well as in collaborative bipartite graphs. In less sparse datasets, a collaborative bipartite graph is usually sufficient. We also highlight that the improvement of recommendation performance through clustering, particularly evident when combined with review embeddings is most visible on sparser data, while on less sparse data incorporating review embeddings may be sufficient when combined with one of the graph-based methods, or otherwise when combined with clustering in other methods.

In this work, we present HMMLVis, an original visualization tool for multivariate Granger causal inference. More precisely, for heterogeneous Granger causality to infer causal relationships in time-series following an exponential distribution. HMMLVis is easy to use and can be applied in any scientific discipline exploring time series and their relationships. In this paper, we focus on climatological and meteorological applications. The visualization tool is demonstrated on different types of applications related to meteorological events on the upper/lower tails of the respective distributions using a renewable energy (wind, PV), air pollution, and the EUMETNET postprocessing benchmark data set (EUPPBench) and different temporal horizons. We demonstrate that the HMMLVis method and visualization depicts the known causal and detects causal relations in the temporal dependencies which are additional important information for the respective cases. We believe that HMMVis as an interpretable visualization tool will serve climatologists or meteorologists and in this way it will contribute to knowledge discovery in these scientific fields.

Spectral clustering is a leading clustering method. Two of its major shortcomings are the disjoint optimization process and the limited representation capacity. To address these issues, we propose a deep spectral clustering model (named BootSC), which jointly learns all stages of spectral clustering—affinity matrix construction, spectral embedding, and $k$-means clustering—using a single network in an end-to-end manner. BootSC leverages effective and efficient optimal-transport-derived supervision to bootstrap the affinity matrix and the cluster assignment matrix. Moreover, a semantically-consistent orthogonal re-parameterization technique is introduced to orthogonalize spectral embeddings, significantly enhancing the discrimination capability. Experimental results indicate that BootSC achieves state-of-the-art clustering performance. For example, it accomplishes a notable 16% NMI improvement over the runner-up method on the challenging ImageNet-Dogs dataset. Our code is available at https://github.com/spdj2271/BootSC.

Density-based clustering algorithms like DBSCAN are highly effective but sensitive to parameter selection, particularly the neighborhood radius (ε) and the minimum number of neighboring points to form a cluster (minPts). We analyze and investigate the influence of the parameter settings onto the clustering outcome under the lense of persistent homology, a technique from topological data analysis. Persistent homology analyzes topological features, such as connected components and loops, across multiple spatial scales, improving clustering accuracy and robustness. We use the density-connectivity distance, a recent finding in the field, to allow full automatization of our approach. In extensive experiments, we demonstrate how insights from persistent homology can help to identify optimal parameter values and introduce an approach to automate parameter selection for density-based clustering. The proposed technique allows DBSCAN and related algorithms to perform effectively on a large variety of datasets without any user input. It combines topological insights with clustering techniques to provide a foundation for robust, automated approaches to complex data analysis.

Detecting anomalies in general ledger data is of utmost importance to ensure the trustworthiness of financial records. Financial audits increasingly rely on machine learning (ML) algorithms to identify irregular or potentially fraudulent journal entries, each characterized by a varying number of transactions. In machine learning, heterogeneity in feature dimensions adds significant complexity to data analysis. In this paper, we introduce a novel approach to anomaly detection in financial data using Large Language Model (LLM) embeddings. To encode non-semantic categorical data (i.e., attributes lacking inherent linguistic meaning) from real-world financial records, we tested 3 pretrained general-purpose sentence-transformer models. For the downstream classification task, we implemented and evaluated 5 optimized ML models, including Logistic Regression, Random Forest, Gradient Boosting Machines, Support Vector Machines, and Neural Networks. Our experiments demonstrate that LLMs contribute valuable information to anomaly detection as our models outperform the baselines, in selected settings by a large margin. The findings further underscore the effectiveness of LLMs in enhancing anomaly detection in financial journal entries, particularly by tackling feature sparsity. We discuss a promising perspective on using SBERT embeddings for non-semantic data in the financial context and beyond.

Although Graph Neural Networks (GNNs) have exhibited the powerful ability to gather graph-structured information from neighborhood nodes via various message-passing mechanisms, the performance of GNNs is limited by poor generalization and fragile robustness caused by noisy and redundant graph data. As a prominent solution, Graph Augmentation Learning (GAL) has recently received increasing attention in the literature. Among the existing GAL approaches, edge-dropping methods that randomly remove edges from a graph during training are effective techniques to improve the robustness of GNNs. However, randomly dropping edges often results in bypassing critical edges. Consequently, the effectiveness of message passing is weakened. In this paper, we propose a novel adversarial edge-dropping method (ADEdgeDrop) that leverages an adversarial edge predictor guiding the removal of edges, which can be flexibly incorporated into diverse GNN backbones. Employing an adversarial training framework, the edge predictor utilizes the line graph transformed from the original graph to estimate the edges to be dropped, which improves the interpretability of the edge-dropping method. The proposed ADEdgeDrop is optimized alternately by stochastic gradient descent and projected gradient descent. Comprehensive experiments on eight graph benchmark datasets demonstrate that the proposed ADEdgeDrop outperforms state-of-the-art baselines across various GNN backbones, demonstrating improved generalization and robustness.