The everyday consumer is inundated with applications powered by machine learning. But in an ordinary day, do we encounter situations and choices which could also benefit from machine learning for which there is no specific tool invented yet? We describe scenarios where people without any machine learning background could find it useful to define their own solution which uses machine learning. Although machine learning is becoming ubiquitous, the average person is unaware of the steps involved. This abstraction makes sense, in many situations, such as traffic predictions, it is not necessary for the driver to know what machine learning algorithm is running. However, we consider examples where knowing how to incorporate machine learning into a problem would assist in decision making. We propose a workflow with operations leading to a final application. There are several challenges here, namely, the average consumer is not expected to have a mathematical background, nor is expected to acquire any additional background. To achieve this new utility, we use a visual analytic pipeline which integrates machine learning and the person. We use the IEEE VAST 2018 Challenge as a case study in which the user steps through the workflow. Finally, we envision the resulting application.

A significant number of soldiers have come forward to report discomfort, irritation, and respiratory problems after taking part in live firing sessions. These problems appear to be caused due to the fumes and particulates emitted from the gun upon firing. There exists substantial research focused on lead and other harmful metallic particulates expelled from a firearm as those are considered the most harmful among the emissions. However, our research focuses on visualizing the carbon monoxide (CO) particles released from a firearm to improve our understanding of their adverse effects on the human body. We use data provided by researchers at Wright-Patterson Air Force Base (WPAFB) enhanced with analyses of provided video material to devise a visualization that shows the correlations between the concentration of CO particles and Lung Deposited Surface Area (LDSA) values as well as other relevant parameters. The results are summarized in the form of a dynamic parallel coordinates plots for in-depth analysis by the domain specialists. Results of this study may be used to glean information about the interrelation between CO particles released and health issues faced by individuals after firing a weapon during a training exercise.

Social media has played a huge role in the 2016 US Presidential Elections. In this paper, we present the results of an exploratory visualization of 200,000+ tweets from confirmed fake twitter accounts. We analyze the user accounts by examining their user names, descriptions/bios, tweets, tweet frequency, and content. We found that they made themselves relatable using political and religious beliefs and then used their influence by joining into popular hashtags on Twitter and posting strongly polarizing tweets at crucial times such as debates and primaries in the election cycle.

We propose novel deep learning based chemometric data analysis technique. We trained L2 regularized sparse autoencoder end-to-end for reducing the size of the feature vector to handle the classic problem of the curse of dimensionality in chemometric data analysis. We introduce a novel technique of automatic selection of nodes inside the hidden layer of an autoencoder through Pareto optimization. Moreover, Gaussian process regressor is applied on the reduced size feature vector for the regression. We evaluated our technique on orange juice and wine dataset and results are compared against 3 state-of-the-art methods. Quantitative results are shown on Normalized Mean Square Error (NMSE) and the results show considerable improvement in the state-of-the-art.

Many research and development activities for scientific data analysis have focused on scalability challenges and data-driven features. Conversely, data visualization that focuses on models requiring human interaction rarely involve practical and largescale scientific data analysis. Therefore, a gap exists between interactive data visualization and scientific data analysis applications. In this paper, we present a design study of interactive data visualization to support scientists who visually analyze data from neutron scattering experiments. This study was conducted in multiple phases: 1) problem characterization; 2) initial design and formative evaluation; and 3) iterative design. We characterize the problems and the design requirements for the analysis of the specific physical science data. We discuss the design, development, evaluation of our visual analytics tool and as well as our iterative developments with physical scientists. We show how to bridge the gap between the two disciplines uncovering new potential to solve their challenges in this design study. We focus on a specific challenge, finding an optimal color mapping, which plays a critical role in neutron scattering science and is broadly applicable to other scientific domains. To address the challenge, we propose two interactive visualization techniques: a dynamic color scale bar (DCSB) and a multi-scale histogram (MSH)

As more and more college classrooms utilize online platforms to facilitate teaching and learning activities, analyzing student online behaviors becomes increasingly important for instructors to effectively monitor and manage student progress and performance. In this paper, we present CCVis, a visual analytics tool for analyzing the course clickstream data and exploring student online learning behaviors. Targeting a large college introductory course with over two thousand student enrollments, our goal is to investigate student behavior patterns and discover the possible relationships between student clickstream behaviors and their course performance. We employ higher-order network and structural identity classification to enable visual analytics of behavior patterns from the massive clickstream data. CCVis includes four coordinated views (the behavior pattern, behavior breakdown, clickstream comparative, and grade distribution views) for user interaction and exploration. We demonstrate the effectiveness of CCVis through case studies along with an ad-hoc expert evaluation. Finally, we discuss the limitation and extension of this work.

Sleep plays an important role in the overall health and wellbeing of a child. The relationship between sleep and daytime behaviours of children with neurodevelopmental disorders is understood poorly; different aspects of a child’s routine may interact with each other to contribute to sleep disorders. In order to diagnose, monitor and successfully treat many medical conditions pertaining to sleep, it becomes imperative to analyse the many aspects of a child’s daytime and sleep behaviours. In this paper, we propose a visual analytic tool for studying the interaction of different variables collected for a child with neurodevelopmental disorders. We propose a visual analytic tool which allows clinicians to explore how the different aspects of a child’s behaviour and activities affect their sleep. This tool is developed as an extension of an existing tool SWAPP, which allows caregivers and clinicians to log and monitor the child’s everyday data.

We present a computer interface to visualize and interact with mathematical knots, i.e., the embeddings of closed circles in 3-dimensional Euclidean space. Mathematical knots are slightly different than everyday knots in that they are infinitely stretchy and flexible when being deformed into their topological equivalence. In this work, we design a visualization interface to depict mathematical knots as closed node-link diagrams with energies charged at each node, so that highly-tangled knots can evolve by themselves from high-energy states to minimal (or lower) energy states. With a family of interactive methods and supplementary user interface elements, out tool allows one to sketch, edit, and experiment with mathematical knots, and observe their topological evolution towards optimal embeddings. In addition, out interface can extract from the entire knot evolution those key moments where successive terms in the sequence differ by critical change; this provides a clear and intuitive way to understand and trace mathematical evolution with a minimal number of visual frames. Finally out interface is adapted and extended to support the depiction of mathematical links and braids, whose mathematical concepts and interactions are just similar to our intuition about knots. All these combine to show a mathematically rich interface to help us explore and understand a family of fundamental geometric and topological problems.

We present VideoSwarm, a system for visualizing video ensembles generated by numerical simulations. VideoSwarm is a web application, where linked views of the ensemble each represent the data using a different level of abstraction. VideoSwarm uses multidimensional scaling to reveal relationships between a set of simulations relative to a single moment in time, and to show the evolution of video similarities over a span of time. VideoSwarm is a plug-in for Slycat, a web-based visualization framework which provides a web-server, database, and Python infrastructure. The Slycat framework provides support for managing multiple users, maintains access control, and requires only a Slycat supported commodity browser (such as Firefox, Chrome, or Safari).