arXiv subjects

Andreas Kerren

Publications and source records attributed to Andreas Kerren.

At least 19 recordsLinked to original sources

arXiv2026

MAPLE: Self-Supervised Learning-Enhanced Nonlinear Dimensionality Reduction for Visual Analysis

We present a new nonlinear dimensionality reduction method, MAPLE, that enhances UMAP by improving manifold modeling. MAPLE employs a self-supervised learning approach to more efficiently encode low-dimensional manifold geometry. Central to this approach are maximum manifold capacity representations (MMCRs), which help untangle complex manifolds by compressing variances among locally similar data points while amplifying variance among dissimilar data points. This design is particularly effective for high-dimensional data with substantial intra-cluster variance and curved manifold structures, such as biological or image data. Our qualitative and quantitative evaluations demonstrate that MAPLE can produce clearer visual cluster separations and finer subcluster resolution than UMAP while maintaining tractable computational cost.

cs.LG↗

arXiv2024

Adversarial Attacks on Machine Learning-Aided Visualizations

Research in ML4VIS investigates how to use machine learning (ML) techniques to generate visualizations, and the field is rapidly growing with high societal impact. However, as with any computational pipeline that employs ML processes, ML4VIS approaches are susceptible to a range of ML-specific adversarial attacks. These attacks can manipulate visualization generations, causing analysts to be tricked and their judgments to be impaired. Due to a lack of synthesis from both visualization and ML perspectives, this security aspect is largely overlooked by the current ML4VIS literature. To bridge this gap, we investigate the potential vulnerabilities of ML-aided visualizations from adversarial attacks using a holistic lens of both visualization and ML perspectives. We first identify the attack surface (i.e., attack entry points) that is unique in ML-aided visualizations. We then exemplify five different adversarial attacks. These examples highlight the range of possible attacks when considering the attack surface and multiple different adversary capabilities. Our results show that adversaries can induce various attacks, such as creating arbitrary and deceptive visualizations, by systematically identifying input attributes that are influential in ML inferences. Based on our observations of the attack surface characteristics and the attack examples, we underline the importance of comprehensive studies of security issues and defense mechanisms as a call of urgency for the ML4VIS community.

cs.CR↗

arXiv2024

Simbanex: Similarity-based Exploration of IEEE VIS Publications

Embeddings are powerful tools for transforming complex and unstructured data into numeric formats suitable for computational analysis tasks. In this work, we use multiple embeddings for similarity calculations to be applied in bibliometrics and scientometrics. We build a multivariate network (MVN) from a large set of scientific publications and explore an aspect-driven analysis approach to reveal similarity patterns in the given publication data. By dividing our MVN into separately embeddable aspects, we are able to obtain a flexible vector representation which we use as input to a novel method of similarity-based clustering. Based on these preprocessing steps, we developed a visual analytics application, called Simbanex, that has been designed for the interactive visual exploration of similarity patterns within the underlying publications.

cs.DL↗

arXiv2024

A Shape Change Enhancing Hierarchical Layout for the Pairwise Comparison of Directed Acyclic Graphs

Comparing directed acyclic graphs is essential in various fields such as healthcare, social media, finance, biology, and marketing. DAGs often result from contagion processes over networks, including information spreading, retweet activity, disease transmission, financial crisis propagation, malware spread, and gene mutations. For instance, in disease spreading, an infected patient can transmit the disease to contacts, making it crucial to analyze and predict scenarios. Similarly, in finance, understanding the effects of saving or not saving specific banks during a crisis is vital. Experts often need to identify small differences between DAGs, such as changes in a few nodes or edges. Even the presence or absence of a single edge can be significant. Visualization plays a crucial role in facilitating these comparisons. However, standard hierarchical layout algorithms struggle to visualize subtle changes effectively. The typical hierarchical layout, with the root on top, is preferred due to its performance in comparison to other layouts. Nevertheless, these standard algorithms prioritize single-graph aesthetics over comparison suitability, making it challenging for users to spot changes. To address this issue, we propose a layout that enhances shape changes in DAGs while minimizing the impact on aesthetics. Our approach involves outwardly swapping changes, altering the DAG's shape. We introduce new drawing criteria. Our layout builds upon a Sugiyama-like hierarchical layout and implements these criteria through two extensions. We designed it this way to maintain interchangeability and accommodate future optimizations, such as pseudo-nodes for edge crossing minimization. In our evaluations, our layout achieves excellent results, with edge crossing aesthetics averaging around 0.8 (on a scale of 0 to 1). Additionally, our layout outperforms the base implementation by an average of 60-75\%.

cs.HC↗

arXiv2024

MetaStackVis: Visually-Assisted Performance Evaluation of Metamodels

Stacking (or stacked generalization) is an ensemble learning method with one main distinctiveness from the rest: even though several base models are trained on the original data set, their predictions are further used as input data for one or more metamodels arranged in at least one extra layer. Composing a stack of models can produce high-performance outcomes, but it usually involves a trial-and-error process. Therefore, our previously developed visual analytics system, StackGenVis, was mainly designed to assist users in choosing a set of top-performing and diverse models by measuring their predictive performance. However, it only employs a single logistic regression metamodel. In this paper, we investigate the impact of alternative metamodels on the performance of stacking ensembles using a novel visualization tool, called MetaStackVis. Our interactive tool helps users to visually explore different singular and pairs of metamodels according to their predictive probabilities and multiple validation metrics, as well as their ability to predict specific problematic data instances. MetaStackVis was evaluated with a usage scenario based on a medical data set and via expert interviews.

cs.LG↗

arXiv2024

Visualization for Trust in Machine Learning Revisited: The State of the Field in 2023

Visualization for explainable and trustworthy machine learning remains one of the most important and heavily researched fields within information visualization and visual analytics with various application domains, such as medicine, finance, and bioinformatics. After our 2020 state-of-the-art report comprising 200 techniques, we have persistently collected peer-reviewed articles describing visualization techniques, categorized them based on the previously established categorization schema consisting of 119 categories, and provided the resulting collection of 542 techniques in an online survey browser. In this survey article, we present the updated findings of new analyses of this dataset as of fall 2023 and discuss trends, insights, and eight open challenges for using visualizations in machine learning. Our results corroborate the rapidly growing trend of visualization techniques for increasing trust in machine learning models in the past three years, with visualization found to help improve popular model explainability methods and check new deep learning architectures, for instance.

cs.HC↗

arXiv2024

t-viSNE: Interactive Assessment and Interpretation of t-SNE Projections

t-Distributed Stochastic Neighbor Embedding (t-SNE) for the visualization of multidimensional data has proven to be a popular approach, with successful applications in a wide range of domains. Despite their usefulness, t-SNE projections can be hard to interpret or even misleading, which hurts the trustworthiness of the results. Understanding the details of t-SNE itself and the reasons behind specific patterns in its output may be a daunting task, especially for non-experts in dimensionality reduction. In this work, we present t-viSNE, an interactive tool for the visual exploration of t-SNE projections that enables analysts to inspect different aspects of their accuracy and meaning, such as the effects of hyper-parameters, distance and neighborhood preservation, densities and costs of specific neighborhoods, and the correlations between dimensions and visual patterns. We propose a coherent, accessible, and well-integrated collection of different views for the visualization of t-SNE projections. The applicability and usability of t-viSNE are demonstrated through hypothetical usage scenarios with real data sets. Finally, we present the results of a user study where the tool's effectiveness was evaluated. By bringing to light information that would normally be lost after running t-SNE, we hope to support analysts in using t-SNE and making its results better understandable.

cs.LG↗

arXiv2024

StackGenVis: Alignment of Data, Algorithms, and Models for Stacking Ensemble Learning Using Performance Metrics

In machine learning (ML), ensemble methods such as bagging, boosting, and stacking are widely-established approaches that regularly achieve top-notch predictive performance. Stacking (also called "stacked generalization") is an ensemble method that combines heterogeneous base models, arranged in at least one layer, and then employs another metamodel to summarize the predictions of those models. Although it may be a highly-effective approach for increasing the predictive performance of ML, generating a stack of models from scratch can be a cumbersome trial-and-error process. This challenge stems from the enormous space of available solutions, with different sets of data instances and features that could be used for training, several algorithms to choose from, and instantiations of these algorithms using diverse parameters (i.e., models) that perform differently according to various metrics. In this work, we present a knowledge generation model, which supports ensemble learning with the use of visualization, and a visual analytics system for stacked generalization. Our system, StackGenVis, assists users in dynamically adapting performance metrics, managing data instances, selecting the most important features for a given data set, choosing a set of top-performant and diverse algorithms, and measuring the predictive performance. In consequence, our proposed tool helps users to decide between distinct models and to reduce the complexity of the resulting stack by removing overpromising and underperforming models. The applicability and effectiveness of StackGenVis are demonstrated with two use cases: a real-world healthcare data set and a collection of data related to sentiment/stance detection in texts. Finally, the tool has been evaluated through interviews with three ML experts.

cs.LG↗

arXiv2024

FeatureEnVi: Visual Analytics for Feature Engineering Using Stepwise Selection and Semi-Automatic Extraction Approaches

The machine learning (ML) life cycle involves a series of iterative steps, from the effective gathering and preparation of the data, including complex feature engineering processes, to the presentation and improvement of results, with various algorithms to choose from in every step. Feature engineering in particular can be very beneficial for ML, leading to numerous improvements such as boosting the predictive results, decreasing computational times, reducing excessive noise, and increasing the transparency behind the decisions taken during the training. Despite that, while several visual analytics tools exist to monitor and control the different stages of the ML life cycle (especially those related to data and algorithms), feature engineering support remains inadequate. In this paper, we present FeatureEnVi, a visual analytics system specifically designed to assist with the feature engineering process. Our proposed system helps users to choose the most important feature, to transform the original features into powerful alternatives, and to experiment with different feature generation combinations. Additionally, data space slicing allows users to explore the impact of features on both local and global scales. FeatureEnVi utilizes multiple automatic feature selection techniques; furthermore, it visually guides users with statistical evidence about the influence of each feature (or subsets of features). The final outcome is the extraction of heavily engineered features, evaluated by multiple validation metrics. The usefulness and applicability of FeatureEnVi are demonstrated with two use cases and a case study. We also report feedback from interviews with two ML experts and a visualization researcher who assessed the effectiveness of our system.

cs.LG↗

arXiv2024

VisEvol: Visual Analytics to Support Hyperparameter Search through Evolutionary Optimization

During the training phase of machine learning (ML) models, it is usually necessary to configure several hyperparameters. This process is computationally intensive and requires an extensive search to infer the best hyperparameter set for the given problem. The challenge is exacerbated by the fact that most ML models are complex internally, and training involves trial-and-error processes that could remarkably affect the predictive result. Moreover, each hyperparameter of an ML algorithm is potentially intertwined with the others, and changing it might result in unforeseeable impacts on the remaining hyperparameters. Evolutionary optimization is a promising method to try and address those issues. According to this method, performant models are stored, while the remainder are improved through crossover and mutation processes inspired by genetic algorithms. We present VisEvol, a visual analytics tool that supports interactive exploration of hyperparameters and intervention in this evolutionary procedure. In summary, our proposed tool helps the user to generate new models through evolution and eventually explore powerful hyperparameter combinations in diverse regions of the extensive hyperparameter space. The outcome is a voting ensemble (with equal rights) that boosts the final predictive performance. The utility and applicability of VisEvol are demonstrated with two use cases and interviews with ML experts who evaluated the effectiveness of the tool.

cs.LG↗

arXiv2024

HardVis: Visual Analytics to Handle Instance Hardness Using Undersampling and Oversampling Techniques

Despite the tremendous advances in machine learning (ML), training with imbalanced data still poses challenges in many real-world applications. Among a series of diverse techniques to solve this problem, sampling algorithms are regarded as an efficient solution. However, the problem is more fundamental, with many works emphasizing the importance of instance hardness. This issue refers to the significance of managing unsafe or potentially noisy instances that are more likely to be misclassified and serve as the root cause of poor classification performance. This paper introduces HardVis, a visual analytics system designed to handle instance hardness mainly in imbalanced classification scenarios. Our proposed system assists users in visually comparing different distributions of data types, selecting types of instances based on local characteristics that will later be affected by the active sampling method, and validating which suggestions from undersampling or oversampling techniques are beneficial for the ML model. Additionally, rather than uniformly undersampling/oversampling a specific class, we allow users to find and sample easy and difficult to classify training instances from all classes. Users can explore subsets of data from different perspectives to decide all those parameters, while HardVis keeps track of their steps and evaluates the model's predictive performance in a test set separately. The end result is a well-balanced data set that boosts the predictive power of the ML model. The efficacy and effectiveness of HardVis are demonstrated with a hypothetical usage scenario and a use case. Finally, we also look at how useful our system is based on feedback we received from ML experts.

cs.LG↗

arXiv2024

DeforestVis: Behavior Analysis of Machine Learning Models with Surrogate Decision Stumps

As the complexity of machine learning (ML) models increases and their application in different (and critical) domains grows, there is a strong demand for more interpretable and trustworthy ML. A direct, model-agnostic, way to interpret such models is to train surrogate models-such as rule sets and decision trees-that sufficiently approximate the original ones while being simpler and easier-to-explain. Yet, rule sets can become very lengthy, with many if-else statements, and decision tree depth grows rapidly when accurately emulating complex ML models. In such cases, both approaches can fail to meet their core goal-providing users with model interpretability. To tackle this, we propose DeforestVis, a visual analytics tool that offers summarization of the behaviour of complex ML models by providing surrogate decision stumps (one-level decision trees) generated with the Adaptive Boosting (AdaBoost) technique. DeforestVis helps users to explore the complexity versus fidelity trade-off by incrementally generating more stumps, creating attribute-based explanations with weighted stumps to justify decision making, and analysing the impact of rule overriding on training instance allocation between one or more stumps. An independent test set allows users to monitor the effectiveness of manual rule changes and form hypotheses based on case-by-case analyses. We show the applicability and usefulness of DeforestVis with two use cases and expert interviews with data analysts and model developers.

cs.LG↗

arXiv2024

VisRuler: Visual Analytics for Extracting Decision Rules from Bagged and Boosted Decision Trees

Bagging and boosting are two popular ensemble methods in machine learning (ML) that produce many individual decision trees. Due to the inherent ensemble characteristic of these methods, they typically outperform single decision trees or other ML models in predictive performance. However, numerous decision paths are generated for each decision tree, increasing the overall complexity of the model and hindering its use in domains that require trustworthy and explainable decisions, such as finance, social care, and health care. Thus, the interpretability of bagging and boosting algorithms, such as random forest and adaptive boosting, reduces as the number of decisions rises. In this paper, we propose a visual analytics tool that aims to assist users in extracting decisions from such ML models via a thorough visual inspection workflow that includes selecting a set of robust and diverse models (originating from different ensemble learning algorithms), choosing important features according to their global contribution, and deciding which decisions are essential for global explanation (or locally, for specific cases). The outcome is a final decision based on the class agreement of several models and the explored manual decisions exported by users. We evaluated the applicability and effectiveness of VisRuler via a use case, a usage scenario, and a user study. The evaluation revealed that most users managed to successfully use our system to explore decision rules visually, performing the proposed tasks and answering the given questions in a satisfying way.

cs.LG↗

arXiv2024

Designing a 3D Gestural Interface to Support User Interaction with Time-Oriented Data as Immersive 3D Radar Chart

The design of intuitive three-dimensional user interfaces is vital for interaction in virtual reality, allowing to effectively close the loop between a human user and the virtual environment. The utilization of 3D gestural input allows for useful hand interaction with virtual content by directly grasping visible objects, or through invisible gestural commands that are associated with corresponding features in the immersive 3D space. The design of such interfaces remains complex and challenging. In this article, we present a design approach for a three-dimensional user interface using 3D gestural input with the aim to facilitate user interaction within the context of Immersive Analytics. Based on a scenario of exploring time-oriented data in immersive virtual reality using 3D Radar Charts, we implemented a rich set of features that is closely aligned with relevant 3D interaction techniques, data analysis tasks, and aspects of hand posture comfort. We conducted an empirical evaluation (n=12), featuring a series of representative tasks to evaluate the developed user interface design prototype. The results, based on questionnaires, observations, and interviews, indicate good usability and an engaging user experience. We are able to reflect on the implemented hand-based grasping and gestural command techniques, identifying aspects for improvement in regard to hand detection and precision as well as emphasizing a prototype's ability to infer user intent for better prevention of unintentional gestures.

cs.HC↗

arXiv2023

2D, 2.5D, or 3D? An Exploratory Study on Multilayer Network Visualisations in Virtual Reality

Relational information between different types of entities is often modelled by a multilayer network (MLN) -- a network with subnetworks represented by layers. The layers of an MLN can be arranged in different ways in a visual representation, however, the impact of the arrangement on the readability of the network is an open question. Therefore, we studied this impact for several commonly occurring tasks related to MLN analysis. Additionally, layer arrangements with a dimensionality beyond 2D, which are common in this scenario, motivate the use of stereoscopic displays. We ran a human subject study utilising a Virtual Reality headset to evaluate 2D, 2.5D, and 3D layer arrangements. The study employs six analysis tasks that cover the spectrum of an MLN task taxonomy, from path finding and pattern identification to comparisons between and across layers. We found no clear overall winner. However, we explore the task-to-arrangement space and derive empirical-based recommendations on the effective use of 2D, 2.5D, and 3D layer arrangements for MLNs.

cs.HC↗

arXiv2023

User Preferences of Spatio-Temporal Referencing Approaches For Immersive 3D Radar Charts

The use of head-mounted display technologies for virtual reality experiences is inherently single-user-centred, allowing for the visual immersion of its user in the computer-generated environment. This isolates them from their physical surroundings, effectively preventing external visual information cues, such as the pointing and referral to an artifact by another user. However, such input is important and desired in collaborative scenarios when exploring and analyzing data in virtual environments together with a peer. In this article, we investigate different designs for making spatio-temporal references, i.e., visually highlighting virtual data artifacts, within the context of Collaborative Immersive Analytics. The ability to make references to data is foundational for collaboration, affecting aspects such as awareness, attention, and common ground. Based on three design options, we implemented a variety of approaches to make spatial and temporal references in an immersive virtual reality environment that featured abstract visualization of spatio-temporal data as 3D Radar Charts. We conducted a user study (n=12) to empirically evaluate aspects such as aesthetic appeal, legibility, and general user preference. The results indicate a unified favour for the presented location approach as a spatial reference while revealing trends towards a preference of mixed temporal reference approaches dependent on the task configuration: pointer for elementary, and outline for synoptic references. Based on immersive data visualization complexity as well as task reference configuration, we argue that it can be beneficial to explore multiple reference approaches as collaborative information cues, as opposed to following a rather uniform user interface design.

cs.HC↗

arXiv2022

Designing Feature Vector Representations: A case study from Chemistry

We present a case study investigating feature descriptors in the context of the analysis of chemical multivariate ensemble data. The data of each ensemble member consists of three parts: the design parameters for each ensemble member, field data resulting from the numerical simulations, and physical properties of the molecules. Since feature-based methods have the potential to reduce the data complexity and facilitate comparison and clustering, we are focusing on such methods. However, there are many options to design the feature vector representation and there is no obvious preference. To get a better understanding of the different representations, we analyze their similarities and differences. Thereby, we focus on three characteristics derived from the representations: the distribution of pairwise distances, the clustering tendency, and the rank-order of the pairwise distances. The results of our investigations partially confirmed expected behavior, but also provided some surprising observations that can be used for the future development of feature representations in the chemical domain.

cs.HC↗

arXiv2020

Xtreaming: an incremental multidimensional projection technique and its application to streaming data

Streaming data applications are becoming more common due to the ability of different information sources to continuously capture or produce data, such as sensors and social media. Despite recent advances, most visualization approaches, in particular, multidimensional projection or dimensionality reduction techniques, cannot be directly applied in such scenarios due to the transient nature of streaming data. Currently, only a few methods address this limitation using online or incremental strategies, continuously processing data, and updating the visualization. Despite their relative success, most of them impose the need for storing and accessing the data multiple times, not being appropriate for streaming where data continuously grow. Others do not impose such requirements but are not capable of updating the position of the data already projected, potentially resulting in visual artifacts. In this paper, we present Xtreaming, a novel incremental projection technique that continuously updates the visual representation to reflect new emerging structures or patterns without visiting the multidimensional data more than once. Our tests show that Xtreaming is competitive in terms of global distance preservation if compared to other streaming and incremental techniques, but it is orders of magnitude faster. To the best of our knowledge, it is the first methodology that is capable of evolving a projection to faithfully represent new emerging structures without the need to store all data, providing reliable results for efficiently and effectively projecting streaming data.

eess.SP↗