We propose a visual analytics approach for the exploration and analysis of dynamic networks. consider snapshots network as points in high-dimensional space project these to two dimensions visualization interaction using juxtaposed views: one showing snapshot evolution network. With this users are enabled detect stable states, recurring outlier topologies, gain knowledge about transitions between states general. The components our discretization, vectorization normalization, dimensionality...

We present a system for the interactive construction and analysis of decision trees that enables domain experts to bring in specific knowledge. identify different user tasks corresponding requirements, develop incorporating tight integration visualization, interaction algorithmic support. Domain are supported growing, pruning, optimizing analysing trees. Furthermore, we scalable tree visualization optimized exploration. show effectiveness our approach by applying methods two use cases. The...

Network data is ubiquitous; e-mail traffic between persons, telecommunication, transport and financial networks are some examples. Often these large multivariate, besides the topological structure of network, multivariate on nodes links available. Currently, exploration analysis methods focused a single aspect; network topology or data. In addition, tools techniques highly domain specific require expert knowledge. We focus non-expert user propose novel solution for that tightly couples...

Abstract We present a novel visual exploration method based on small multiples and large singles for effective efficient data analysis. Users are enabled to explore the state space by offering multiple alternatives from current state. can then select alternative of choice continue Furthermore, intermediate steps in process preserved be revisited adapted using an intuitive navigation mechanism well‐known undo‐redo stack filmstrip metaphor. As proof concept is implemented prototype. The...

Networks are present in many fields such as finance, sociology, and transportation. Often these networks dynamic: they have a structural well temporal aspect. In addition to relations occurring over time, node information is frequently hierarchical structure or time-series data. We technique that extends the Massive Sequence View ( msv) for analysis of aspects dynamic networks. Using features data Gestalt principles visualization closure, proximity, similarity, we developed reordering...

Networks are present in many fields such as finance, sociology, and transportation. Often these networks dynamic: they have a structural well temporal aspect. We technique that extends the Massive Sequence View (MSV) for analysis of aspects dynamic networks. Using features data visualization based on Gestalt principles closure, proximity, similarity, we developed node reordering strategies MSV to make stand out. This enables users find properties trends, counter periodicity, shifts,...

We present a conceptual framework for the development of visual interactive techniques to formalize and externalize trust in machine learning (ML) workflows. Currently, ML applications is an implicit process that takes place user's mind. As such, there no method feedback or communication can be acted upon. Our will instrumental developing visualization approaches help users efficiently effectively build communicate ways fit each stages. formulate several research questions directions...

Data features and class probabilities are two main perspectives when, e.g., evaluating model results identifying problematic items. Class represent the likelihood that each instance belongs to a particular class, which can be produced by probabilistic classifiers or even human labeling with uncertainty. Since both multi-dimensional data, dimensionality reduction (DR) techniques commonly used extract informative characteristics from them. However, existing methods either focus solely on data...

Abstract Model comparison is an important process to facilitate model diagnosis, improvement, and selection when multiple models are developed for a classification task. It involves careful concerning performance interpretation. Current visual analytics solutions often ignore the feature process. They either do not support detailed analysis of multi‐class classifiers or rely on alone interpret results. Understanding how different make decisions, especially disagreements same instances,...

Event sequence data is a special type of time-dependent that captures information about the order in which discrete events occur. The time-dimension one factors makes event hard to understand. Other contribute this complexity are dimensionality terms amount and attributes, frequency consistency density, varying durations, parallel occurrences. When end-users need compare sequences, all these characteristics be considered justified. In state-of-the-art report we review visualization...

Abstract Event sequences and time series are widely recorded in many application domains; examples stock market prices, electronic health records, server operation performance logs. Common goals for recording monitoring, root cause analysis predictive analytics. Current methods generally focus on the exploration of either event or series. However, deeper insights gained by combining both. We present a visual analytics approach where users can explore both data simultaneously, visualization,...

<p>Citation Information: V. Prasad, R. van Sloun, S. den Elzen, A. Vilanova, and N. Pezzotti, “The Transform-and-Perform framework: Explainable deep learning beyond classification,” IEEE Transactions on Visualization Computer Graphics, 2022. https://doi.org/10.1109/TVCG.2022.3219248</p> <p><br></p> <p>© 2022 IEEE. Personal use of this material is permitted. Permission from must be obtained for all other uses, in any current or future media, including...

Hypnograms contain a wealth of information and play an important role in sleep medicine. However, interpretation the hypnogram is difficult task requires domain knowledge "clinical intuition." This study aimed to uncover which features drive by physicians. In other words, make explicit physicians implicitly look for hypnograms. Three experts evaluated up 612 hypnograms, indicating normal or abnormal structure suspicion disorders. ElasticNet convolutional neural network classification models...

In the vast landscape of visualization research, Dimensionality Reduction (DR) and graph analysis are two popular subfields, often essential to most visual data analytics setups. DR aims create representations support neighborhood similarity on complex, large datasets. Graph focuses identifying salient topological properties key actors within networked data, with specialized research investigating how such features could be presented user ease comprehension underlying structure. Although...

In personalized recommender systems, embeddings are often used to encode customer actions and items, retrieval is then performed in the embedding space using approximate nearest neighbor search. However, this approach can lead two challenges: 1) user restrict diversity of interests captured 2) need keep them up-to-date requires an expensive, real-time infrastructure. paper, we propose a method that overcomes these challenges practical, industrial setting. The dynamically updates profiles...

In recent years, visual analytics (VA) has shown promise in alleviating the challenges of interpreting black-box deep learning (DL) models. While focus VA for explainable DL been mainly on classification problems, is gaining popularity high-dimensional-to-high-dimensional ( <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">H-H</i> ) problems such as image-to-image translation. contrast to classification, have no explicit instance groups or...

&lt;p&gt;In recent years, visual analytics (VA) has shown promise in alleviating the challenges of interpreting black-box deep learning (DL) models. While focus VA for explainable DL been mainly on classification problems, is gaining popularity high-dimensional-to-high-dimensional (H-H) problems such as image-to-image translation. In contrast to classification, H-H have no explicit instance groups or classes study. Each output continuous, high dimensional, and changes an unknown non-linear...

Abstract In many domains, multivariate event sequence data is collected focused around an entity (the case). Typically, each has multiple attributes, for example, in healthcare a patient events such as hospitalization, medication, and surgery. addition to the events, also case (a specific attribute, e.g., patient) associated (e.g., age, gender, weight). Current work typically only visualizes one attribute per (label) sequences. As consequence, can be explored from predefined case‐centric...

Data features and class probabilities are two main perspectives when, e.g., evaluating model results identifying problematic items. Class represent the likelihood that each instance belongs to a particular class, which can be produced by probabilistic classifiers or even human labeling with uncertainty. Since both multi-dimensional data, dimensionality reduction (DR) techniques commonly used extract informative characteristics from them. However, existing methods either focus solely on data...

The group dynamics visualizer is a solution to the problem healthcare workers are facing when working with social therapeutic groups. It well known that their reflection ability affected negatively by dynamics. Healthcare become part of this groups and no longer able objectively observe intervene. In paper we describe process measuring dynamics, 'group prevention plan'; more specific, giving feedback on tension level inside psychiatric ward. A software application, visualizer', developed...

&lt;p&gt;Citation Information: V. Prasad, R. van Sloun, S. den Elzen, A. Vilanova, and N. Pezzotti, “The Transform-and-Perform framework: Explainable deep learning beyond classification,” IEEE Transactions on Visualization Computer Graphics, 2022. https://doi.org/10.1109/TVCG.2022.3219248&lt;/p&gt; &lt;p&gt;&lt;br&gt;&lt;/p&gt; &lt;p&gt;© 2022 IEEE. Personal use of this material is permitted. Permission from must be obtained for all other uses, in any current or future media, including...