All Courses
-
JBI100 (2017-4) VisualizationVisualization
In the visualization course you will learn about the challenges of visually representing data that comes in a variety of forms. Starting from simple primitive data types like categorical, ordinal, or quantitative data, we will have a look into more complex dataset scenarios including relational data like graphs/networks or hierarchies, multivariate data, text data, or trajectory data that contains an inherent spatio-temporal aspect. Moreover, as with trajectory data, all of the discussed data types can have a time-varying nature, requiring even more sophisticated visualization techniques. A dataset might also be a mixture of several data types making it even more challenging to build a powerful visualization with the goal to find insights in the data. In this course you will learn about the data processing, data transformation, data visualization, and finally, the interaction with the visual output. As you can see, visualization has to tackle different stages in order to be useful as a means for data analysis. It is not just producing static aesthetically appealing diagrams. For this reason, you will also learn about algorithmic concepts coming from the field of data science with the goal to preprocess and handle the data to even faster find patterns, correlations, or outliers/anomalies in an unknown dataset. To make a visualization interpretable, readable, and intuitive, we will also have a look at perceptual issues like pre-attentive processing, the visual memory, or Gestalt principles. Moreover, a number of laws or no-goes will be discussed to make the diagrams or visualization techniques more perceptually effective. A combination of many of those concepts like algorithmic transformations of the data, the visualization of the data, and the interaction with the visual output combined with the user-in-the-loop is the general process of typical visual analytics approaches. The goal is to build, refine, reject, or confirm hypotheses with the final goal to detect insights, derive knowledge, and to get some kind of Aha-effect. Visualization is a powerful means in this whole process since it makes use of the perceptual abilities and strengths of the human's visual system to rapidly detect patterns. Visualization makes sense if a totally algorithmic solution will fail due to the fact that the problem cannot be completely specified in terms of input parameters. In visual analytics this interplay of algorithms and interactive visualizations builds a crucial aspect to make it a powerful concept for data analytics. "Computers are incredibly fast, accurate and stupid; humans are incredibly slow, inaccurate and brilliant; together they are powerful beyond imagination." --- Leo Cherne