Parallel Coordinates Research Articles

Context-Aware Visual Abstraction of Crowded Parallel Coordinates

As the size of dataset increases, it has become a difficult task to explore structures of interest from crowded parallel coordinates due to visual clutter. Numerous methods have been proposed to simplify the visualization of crowded parallel coordinates, such as filtering, bundling and sampling. However, contextual structures are hardly preserved in the course of simplification, which make significant features easily lost in the simplified parallel coordinates. In this paper, we propose a context-aware visual sampling method for the exploration of crowded parallel coordinates. A Doc2Vec model, widely used in the field of Natural Language Processing (NLP), is utilized to represent the contextual structures across a series of attribute axes with quantifiable vectors. Then, an adaptive blue noise sampling model is employed to reduce the size of original dataset in the vectorized space, guarantying that data items with different contextual structures would be retained in the simplified parallel coordinates. A set of meaningful visual interfaces are designed, enabling users to easily capture the contextual features and evaluate the sampled parallel coordinates. Case studies based on real-world datasets and quantitative comparison have demonstrated the effectiveness of our method in the simplification of crowded parallel coordinates and the exploration of large scale multi-dimensional datasets.

Matching in cluster randomized trials using the Goldilocks Approach

In group or cluster-randomized trials (GRTs), matching is a technique that can be used to improve covariate balance. When baseline data are available, we suggest a strategy that can be used to achieve the desired balance between treatment and control groups across numerous potential confounding variables. This strategy minimizes the overall within-pair Mahalanobis distance; and involves iteratively: 1) making pairs that minimize the distance between pairs of clusters with respect to potentially confounding variables; 2) visually assessing the potential effects of these pairs and resulting possible randomizations; and 3) reweighting variables of selecting weights to make pairs of clusters. In step 2, we plot the between-arm differences with a parallel-coordinates plot. Investigators can compare plots of different weighting schemes to determine the one that best suits their needs prior to the actual, final, randomization. We demonstrate application of the approach with the Mupirocin-Iodophor Swap Out trial. A webapp is provided.

Data-driven gait analysis for diagnosis and severity rating of Parkinson’s disease

Parkinsons disease (PD) is the second most neurodegenerative disease, which results in gradual loss of movements. To diagnose PD in a clinical setting, clinicians generally use clinical manifestations like motor and non-motor symptoms and rate the severity based on unified Parkinsons disease rating scale (UPDRS). Such clinical assessment largely depends on the expertise and experience of the clinicians and it is subjective leading to variation in assessment between clinicians. As the gait of people with Parkinson’s generally differs from gait of healthy age-matched adults, the assessment of gait abnormalities can lead to not only the diagnosis of PD but also the rating of severity level based on motor symptoms. Hence, in this paper, a data-driven gait classification framework using the supervised machine learning algorithms is presented. Using the publicly available gait datasets acquired using vertical ground reaction force (VGRF) sensors, we present a correlation based feature extraction technique for improved stage classification of PD. Significant biomarkers from spatiotemporal gait features are obtained based on the correlation, and the normal distribution of the gait dataset is assessed using the Shapiro–Wilk test. Subsequently, four supervised machine learning algorithms, namely, K-nearest neighbours (KNN), Naive Bayes (NB), Ensemble classifier (EC) and Support vector machine (SVM) are used to rate the severity level of PD according to the Hoehn and Yahr (H&Y) scale. The performance of the classifiers, assessed using the confusion matrix and parallel coordinate plots, highlights that SVM can result in a classification accuracy of 98.4%. Moreover, with minimal gait feature set acquired based on the rank correlation, the proposed approach outperforms several other state-of-the-art methods that have used the same dataset for PD stage classification.

Improving Visualization Design for Effective Multi-Objective Decision Making.

Decision-makers across many professions are often required to make multi-objective decisions over increasingly larger volumes of data with several competing criteria. Data visualization is a powerful tool for exploring these complex 'solution spaces', but there is limited research on its ability to support multi-objective decisions. In this article, we explore the effects of chart complexity and data volume on decision quality in multi-objective scenarios with complex trade-offs. We look at the impact of four common multidimensional chart types (scatter plot matrices, parallel coordinates plots, heat maps, radar charts), the number of options and dimensions and participant chart usage experience on decision time and accuracy when selecting the 'optimal option'. As objectively evaluating the quality of multi-objective decisions and the trade-offs involved is challenging, we employ rank- and score-based accuracy metrics. While heat maps demonstrate a time advantage, our findings show no strong performance benefit for one chart type over another for accuracy. We find mixed evidence for the impact of chart complexity on performance, with our results suggesting the existence of a 'ceiling' in the number of dimensions considered by participants. This points to a potential limit to data complexity that is useful for decision making. Lastly, participants who use charts frequently performed better, suggesting that users can potentially be trained to effectively use complex visualizations in their decision-making.

Clustering Algorithms: An Application for Adsorption Kinetic Curves

Clustering algorithms have been used in different areas of knowledge with different goals such as noise detection, outliers, and descriptive tasks. The adsorption kinetics is a curve that describes the rate retention to the adsorbate on the adsorbent at time, which is represents as a two-dimensional graph. In this paper, we present a computational application to determine the experimental conditions that influence when equilibrium point is reached into adsorption kinetics curve using the K-means clustering algorithm and, Parallel Coordinates concept, in order to prove our method we used adsorption kinetic curves Q-PVA . Results obtained were compared with two designs of experiments (three-stage nested design and hierarchical design with crossed factors).

Visual Analysis of Multi-Parameter Distributions Across Ensembles of 3D Fields.

For an ensemble of 3D multi-parameter fields, we present a visual analytics workflow to analyse whether and which parts of a selected multi-parameter distribution is present in all ensemble members. Supported by a parallel coordinate plot, a multi-parameter brush is applied to all ensemble members to select data points with similar multi-parameter distribution. By a combination of spatial sub-division and a covariance analysis of partitioned sub-sets of data points, a tight partition in multi-parameter space with reduced number of selected data points is obtained. To assess the representativeness of the selected multi-parameter distribution across the ensemble, we propose a novel extension of violin plots that can show multiple parameter distributions simultaneously. We investigate the visual design that effectively conveys (dis-)similarities in multi-parameter distributions, and demonstrate that users can quickly comprehend parameter-specific differences regarding distribution shape and representativeness from a side-by-side view of these plots. In a 3D spatial view, users can analyse and compare the spatial distribution of selected data points in different ensemble members via interval-based isosurface raycasting. In two real-world application cases we show how our approach is used to analyse the multi-parameter distributions across an ensemble of 3D fields.

Foveated Encoding for Large High-Resolution Displays

Collaborative exploration of scientific data sets across large high-resolution displays requires both high visual detail as well as low-latency transfer of image data (oftentimes inducing the need to trade one for the other). In this work, we present a system that dynamically adapts the encoding quality in such systems in a way that reduces the required bandwidth without impacting the details perceived by one or more observers. Humans perceive sharp, colourful details, in the small foveal region around the centre of the field of view, while information in the periphery is perceived blurred and colourless. We account for this by tracking the gaze of observers, and respectively adapting the quality parameter of each macroblock used by the H.264 encoder, considering the so-called visual acuity fall-off. This allows to substantially reduce the required bandwidth with barely noticeable changes in visual quality, which is crucial for collaborative analysis across display walls at different locations. We demonstrate the reduced overall required bandwidth and the high quality inside the foveated regions using particle rendering and parallel coordinates.

Visual Analysis of Large Multivariate Scattered Data using Clustering and Probabilistic Summaries.

Rapidly growing data sizes of scientific simulations pose significant challenges for interactive visualization and analysis techniques. In this work, we propose a compact probabilistic representation to interactively visualize large scattered datasets. In contrast to previous approaches that represent blocks of volumetric data using probability distributions, we model clusters of arbitrarily structured multivariate data. In detail, we discuss how to efficiently represent and store a high-dimensional distribution for each cluster. We observe that it suffices to consider low-dimensional marginal distributions for two or three data dimensions at a time to employ common visual analysis techniques. Based on this observation, we represent high-dimensional distributions by combinations of low-dimensional Gaussian mixture models. We discuss the application of common interactive visual analysis techniques to this representation. In particular, we investigate several frequency-based views, such as density plots in 1D and 2D, density-based parallel coordinates, and a time histogram. We visualize the uncertainty introduced by the representation, discuss a level-of-detail mechanism, and explicitly visualize outliers. Furthermore, we propose a spatial visualization by splatting anisotropic 3D Gaussians for which we derive a closed-form solution. Lastly, we describe the application of brushing and linking to this clustered representation. Our evaluation on several large, real-world datasets demonstrates the scaling of our approach.

Uncertainty in Continuous Scatterplots, Continuous Parallel Coordinates, and Fibers.

In this paper, we introduce uncertainty to continuous scatterplots and continuous parallel coordinates. We derive respective models, validate them with sampling-based brute-force schemes, and present acceleration strategies for their computation. At the same time, we show that our approach lends itself as well for introducing uncertainty into the definition of fibers in bivariate data. Finally, we demonstrate the properties and the utility of our approach using specifically designed synthetic cases and simulated data.

Fault severity detection of a worm gearbox based on several feature extraction methods through a developed condition monitoring system

This study presents the severity detection of pitting faults on worm gearbox through the assessment of fault features extracted from the gearbox vibration data. Fault severity assessment on worm gearbox is conducted by the developed condition monitoring instrument with observing not only traditional but also multidisciplinary features. It is well known that the sliding motion between the worm gear and wheel gear causes difficulties about fault detection on worm gearboxes. Therefore, continuous monitoring and observation of different types of fault features are very important, especially for worm gearboxes. Therefore, in this study, time-domain statistics, the features of evaluated vibration analysis method and Poincaré plot are examined for fault severity detection on worm gearbox. The most reliable features for fault detection on worm gearbox are determined via the parallel coordinate plot. The abnormality detection during worm gearbox operation with the developed system is performed successfully by means of a decision tree.

Multivariate statistical process control methods for batch production: a review focused on applications

ABSTRACT In this paper, we highlight the basic techniques of multivariate statistical process control (MSPC) under the dimensionality criteria, such as Multiway Principal Component Analysis, Multiway Partial Squares, Structuration à Trois Indices de la Statistique, Tucker3, Parallel Factors, Multiway Independent Component Analysis, Multiset Canonical Correlation Analysis, Slow Features Analysis, and Parallel Coordinates. Furthermore, we summarize the procedures of each statistical technique and the usage of multivariate control charts. In addition, we review the most significant proposals and applications in practical cases. Finally, we compare and discuss the benefits and limitations within methods.

Using Stakeholder Preferences to Identify Optimal Land Use Configurations

One way to solve multi-objective spatial land use allocation problems is to calculate a set of Pareto-optimal solutions and include stakeholder preferences after the optimization process. There are various land use allocation studies that identify the Pareto frontier (i.e., trade-off curve); to our knowledge, however, for the majority of them, the debate on which solutions are preferred by stakeholders or are preferred by stakeholders remains open. One reason could be that Pareto-optimal solutions, due to their multi-dimensionality, are difficult to communicate. To fill this gap, we give an example using the results of a multi-objective agricultural land use allocation problem that maximizes four biophysical objectives: agricultural production, water quality, water quantity, and biodiversity in the Lossa River Basin in Central Germany. We conducted expert interviews with 11 local stakeholders from different backgrounds, e.g., water experts, nature conservationists, farmers, etc. In addition to providing information about the case study area, we visualized the trade-offs between the different objectives using parallel coordinates plots that allowed the stakeholders to browse through the optimal solutions. Based on this information, the stakeholders set weights for each of the objectives by applying the Analytic Hierarchy Process (AHP). With these weights, we selected the preferred solutions from the Pareto-optimal set. The results show that, overall, stakeholders clearly ranked water quality first, followed by biodiversity, water quantity, and agricultural production. The corresponding land use maps show a huge difference in land management (e.g., less application of fertilizer, more linear elements, and conservation tillage) for the preferred solutions compared to the current status. The method presented in this study can help decision makers finding land use and land management strategies based on both biophysical modeling results and stakeholder expertise, and it shows how multi-objective optimization results can be communicated and used for an information-based decision-making process.

Parallel Coordinates Visualization in the ELK Stack

Modern large-scale distributed computing systems, processing large volumes of data, require mature monitoring systems able to control and track in re-sources, networks, computing tasks, queues and other components. In recent years, the ELK stack has become very popular for the monitoring of computing environment, largely due to the efficiency and flexibility of the Elastic Search storage and wide variety of Kibana visualization tools. The analysis of computing infrastructure metadata often requires the visual exploration of multiple parameters simultaneously on one graphical image. Stacked bar charts, heat maps, radar charts are widely used for the multivariate visual data analysis, but these methods have limitations on the number of parameters. In this research the authors propose to enhance the capacity of Kibana, adding Parallel Coordinates diagram - one of the most powerful method for visual interactive analysis of high-dimensional data. It allows to compare many variables together and observe correlations between them. This work describes the development process of Parallel Coordinates as a Kibana plugin, and demonstrates an example of visual data analysis based on the Nginx logs metadata.

Spinel web: an interactive web application for visualizing the chemical composition of spinel group minerals

The spinel group minerals provide useful information regarding the geological environment in which the host rocks were formed, constituting excellent petrogenetic indicators, and guides in the search for mineral deposits of economic interest. In this article, we present the Spinel Web, a web application to visualize the chemical composition of spinel group minerals. Spinel Web integrates most of the diagrams commonly used for analyzing the chemical characteristics of the spinel group minerals. It incorporates parallel coordinates and a 3D representation of the spinel prisms. It also provides coordinated views and appropriate interactions for users to interact with their datasets. Spinel Web also supports semi-automatic categorization of the geological environment of formation through a standard Web browser.

Augmenting Parallel Coordinates Plots With Color-Coded Stacked Histograms.

We introduce Parallel Histogram Plot (PHP), a technique that overcomes the innate limitations of parallel coordinates plot (PCP) by attaching stacked-bar histograms with discrete color schemes to PCP. The color-coded histograms enable users to see an overview of the whole data without cluttering or scalability issues. Each rectangle in the PHP histograms is color coded according to the data ranking by a selected attribute. This color-coding scheme allows users to visually examine relationships between attributes, even between those that are displayed far apart, without repositioning or reordering axes. We adopt the Visual Information Seeking Mantra so that the polylines of the original PCP can be used to show details of a small number of selected items when the cluttering problem subsides. We also design interactions, such as a focus+context technique, to help users investigate small regions of interest in a space-efficient manner. We provide a real-world example in which PHP is effectively utilized compared with other visualizations, and we perform a controlled user study to evaluate the performance of PHP in helping users estimate the correlation between attributes. The results demonstrate that the performance of PHP was consistent in the estimation of correlations between two attributes regardless of the distance between them.

Multitouch Interaction with Parallel Coordinates on Large Vertical Displays

This paper presents a multitouch vocabulary for interacting with parallel coordinates plots on wall-sized displays. The gesture set relies on principles such as two-finger range definition, a functional distinction of background and foreground for applying the Hold-and-Move concept to wall-sized displays as well as fling-based interaction for triggering and controlling long-range movements. Our implementation demonstrates that out-of-reach problems and limitations regarding multitouch technology and display size can be tackled by the coherent integration of our multitouch gestures. Expert reviews indicate that our gesture vocabulary helps to solve typical analysis tasks that require interaction beyond arms' reach, and it also shows how often certain gestures were used.

Consecutive Compensator in Station-Keeping of a Surface Vessel

This paper addresses the problem of station-keeping of a surface vessel by means of the consecutive compensator approach. The horizontal motion of the vessel is described by a dynamic model. The model is set up in vessel parallel coordinates, with three degrees of freedom: longitudinal, transverse and rotational motion. It is assumed that the vessel is fully actuated, i.e. there is a sufficient number and type of actuators and a thrust allocation system to ensure full manoeuvrability. Thus, the control can be designed with the assumption of three independent inputs and three output signals. The longitudinal motion can be considered separately, but a cross-coupling exists between the transverse and rotational kinetics. There is uncertainty both in parameters and signals, due to the vessel mass, inertia, and damping, as well as the unmeasured derivatives. The proposed control ensures station-keeping when the vessel is subjected to external disturbances. The consecutive compensator, which is based on high-gain feedback, provides robustness. Stability analysis is presented considering the cross-terms as limited disturbances. This allows proof of exponential stability. Experimental results are included from the Marine Cybernetics Laboratory (MC lab) at the Centre for Autonomous Marine Operations and Systems (A MOS) at the Norwegian University of Science and Technology (Norges teknisk-naturvitenskapelige universitet, NTNU ). Two scenarios are investigated: the scaled vessel is subjected to external disturbance, and the vessel executes the " four corner test". The experiments illustrate the applicability of the method.

Comparative study of Mueller matrix transformation and polar decomposition for optical characterization of turbid media

Mueller matrix polarimetry (MMP) is a fast and non-invasive optical technique for characterization of turbid media such as biological tissues, where the experimental Mueller matrix M contains the complete information of polarization properties of the sample. The individual optical properties in M are present in a complex interwoven way and need to be extracted; Mueller matrix polar decomposition (MMPD) and Mueller matrix transformation (MMT) are two widely used methods for this purpose. Here, we compared these two methods (i.e., MMPD and MMT) and assessed the possible correlation between their corresponding parameters in a comprehensive sample cohort containing different types of tissues and associated pathologies (n = 47), tissue phantoms (n = 21) and standard optical components (e.g., air, polarizer, quarter wave plate, etc.) (n = 09). Specifically, we calculated two sets of optical variables (ΔT, b) and (δ, t) representing depolarization and retardance for the MMPD and MMT methods, respectively. Qualitative correlation via graphical tools (i.e., violin, parallel coordinate, scatter, Bland and Altman plots) in tandem with correlation coefficients was investigated. The parameters of both methods are indicative of the structural features of the turbid samples; however, the MMT parameters (b, t) give higher values than the MMPD parameters (ΔT,δ) with moderate statistical correlation. This study assessed the link between the two methods and would provide a useful reference for such comparative analyses.

Visual analytics and prediction system based on deep belief networks for icing monitoring data of overhead power transmission lines

In this paper, a system is proposed for visualizing and analyzing icing monitoring data of power transmission lines. The distributions of temperature and humidity are visualized by two-dimensional maps with customizable map layers. The multi-dimensional monitoring data are visualized as parallel coordinates. Moreover, a prediction algorithm that is based on a hybrid deep belief network is integrated into the system for predicting the icing thickness. If the icing thickness of a certain location exceeds the threshold value, the historical meteorological data of the location can be visualized as line graphs, which helps to choose the appropriate de-icing measures. According to the experimental results, our system is capable of reflecting the statistical features of icing monitoring data with high accuracy of icing thickness prediction.

Shallow Versus Deep Neural Networks in Gear Fault Diagnosis

Accurate gear defect detection in induction machine-based systems is a fundamental issue in several industrial applications. At this aim, shallow neural networks, i.e., architectures with only one hidden layer, have been used after a feature extraction step from vibration, torque, acoustic pressure and electrical signals. Their additional complexity is justified by their ability in extracting its own features and in the very high-test classification rates. These signals are here analyzed, both geometrically and topologically, in order to estimate the class manifolds and their reciprocal positioning. At this aim, the different states of the gears are studied by using linear (Pareto charts, biplots, principal angles) and nonlinear (curvilinear component analysis) techniques, while the class clusters are visualized by using the parallel coordinates. It is deduced that the class manifolds are compact and well separated. This result justifies the use of a shallow neural network, instead of a deep one, as already remarked in the literature, but with no theoretical justification. The experimental section confirms this assertion, and also compares the shallow neural network results with the other machine learning techniques used in the literature.