Search

In this paper, we introduce a visual analytics approach aimed at helping machine learning experts analyze the hidden states of layers in recurrent neural networks. Our technique allows the user to interactively inspect how hidden states store and process information throughout the feeding of an input sequence into the network. The technique can help answer questions, such as which parts of the input data have a higher impact on the prediction and how the model correlates each hidden state configuration with a certain output. Our visual analytics approach comprises several components: First, our input visualization shows the input sequence and how it relates to the output (using color coding). In addition, hidden states are visualized through a nonlinear projection into a 2-D visualization space using t-distributed stochastic neighbor embedding to understand the shape of the space of the hidden states. Trajectories are also employed to show the details of the evolution of the hidden state configurations. Finally, a time-multi-class heatmap matrix visualizes the evolution of the expected predictions for multi-class classifiers, and a histogram indicates the distances between the hidden states within the original space. The different visualizations are shown simultaneously in multiple views and support brushing-and-linking to facilitate the analysis of the classifications and debugging for misclassified input sequences. To demonstrate the capability of our approach, we discuss two typical use cases for long short-term memory models applied to two widely used natural language processing datasets. 3AkgHYPUvew9RocThgM64- Video abstract. [ABSTRACT FROM AUTHOR]

We propose a visualization method to understand the effect of multidimensional projection on local subspaces, using implicit function differentiation. Here, we understand the local subspace as the multidimensional local neighborhood of data points. Existing methods focus on the projection of multidimensional data points, and the neighborhood information is ignored. Our method is able to analyze the shape and directional information of the local subspace to gain more insights into the global structure of the data through the perception of local structures. Local subspaces are fitted by multidimensional ellipses that are spanned by basis vectors. An accurate and efficient vector transformation method is proposed based on analytical differentiation of multidimensional projections formulated as implicit functions. The results are visualized as glyphs and analyzed using a full set of specifically-designed interactions supported in our efficient web-based visualization tool. The usefulness of our method is demonstrated using various multi- and high-dimensional benchmark datasets. Our implicit differentiation vector transformation is evaluated through numerical comparisons; the overall method is evaluated through exploration examples and use cases. [ABSTRACT FROM AUTHOR]

Parallel coordinates provide a flexible and effective way of showing multidimensional data. The authors describe basic concepts, geometric models, and their use in visualization. [ABSTRACT FROM AUTHOR]

The past 10 years have seen profound changes in visualization algorithms, techniques, methodologies, and applications. These changes are forcing alterations to visualization courses. Unfortunately, outdated course content recommendations, together with profound changes in the underlying technology and methodology, are producing an unstable ground for educators at a time when visual representations are becoming increasingly important. To address this issue, educators held meetings or workshops at Siggraph 2011 and 2012 and a panel and workshop at Eurographics 2012. This article presents the insights gathered at these events. [ABSTRACT FROM AUTHOR]

Scatterplots are well established means of visualizing discrete data values with two data variables as a collection of discrete points. We aim at generalizing the concept of scatterplots to the visualization of spatially continuous input data by a continuous and dense plot. An example of a continuous input field is data defined on an n-D spatial grid with respective interpolation or reconstruction of in-between values. We propose a rigorous, accurate, and generic mathematical model of continuous scatterplots that considers an arbitrary density defined on an input field on an n-D domain and that maps this density to m-D scatterplots. Special cases are derived from this generic model and discussed in detail: scatterplots where the n-D spatial domain and the m-D data- attribute domain have identical dimension, 1-D scatterplots as a way to define continuous histograms, and 2-D scatterplots of data on 3-D spatial grids. We show how continuous histograms are related to traditional discrete histograms and to the histograms of isosurface statistics. Based on the mathematical model of continuous scatterplots, respective visualization algorithms are derived, in particular for 2-D scatterplots of data from 3-D tetrahedral grids. For several visualization tasks, we show the applicability of continuous scatterplots. Since continuous scatterplots do not only sample data at grid points but interpolate data values within cells, a dense and complete visualization of the data set is achieved that scales well with increasing data set size. Especially for irregular grids with varying cell size, improved results are obtained when compared to conventional scatterplots. Therefore, continuous scatterplots are a suitable extension of a statistics visualization technique to be applied to typical data from scientific computation. [ABSTRACT FROM AUTHOR]

The article offers information on the 5th Symposium on Biological Data Visualization held in Dublin, Ireland from July 10-11, 2015. The aim of the symposium was to organize a premier international event for all aspects of visualization in biology and was affiliated with ISMB, the Intelligent Systems for Molecular Biology conference, as a Special Interest Group (SIG).

Information about the BioVis 2015 - the fifth Symposium on Biological Data Visualization is presented. Topics discussed include biological data visualization, bioinformatics, and the use of visaul analytics Graphie.

Visual analytics (VA) systems help data analysts solve complex problems interactively, by integrating automated data analysis and mining, such as machine learning (ML) based methods, with interactive visualizations. We propose a conceptual framework that models human interactions with ML components in the VA process, and that puts the central relationship between automated algorithms and interactive visualizations into sharp focus. The framework is illustrated with several examples and we further elaborate on the interactive ML process by identifying key scenarios where ML methods are combined with human feedback through interactive visualization. We derive five open research challenges at the intersection of ML and visualization research, whose solution should lead to more effective data analysis. [ABSTRACT FROM AUTHOR]