China 
Africa 
Explore chapters and articles related to this topic
Model Development
Published in Przemyslaw Biecek, Tomasz Burzykowski, Explanatory Model Analysis, 2021
Przemyslaw Biecek, Tomasz Burzykowski
In this book, we rely on five visualization techniques for data exploration, schematically presented in Figure 2.3. Two of them (histogram and empirical cumulative-distribution (ECD) plot) are used to summarize the distribution of a single random (explanatory or dependent) variable; the remaining three (mosaic plot, box plot, and scatter plot) are used to explore the relationship between pairs of variables. Note that a histogram can be used to explore the distribution of a continuous or a categorical variable, while ECD and box plots are suitable for continuous variables. A mosaic plot is useful for exploring the relationship between two categorical variables, while a scatter plot can be applied for two continuous variables. It is worth noting that box plots can also be used for evaluating a relation between a categorical variable and a continuous one, as illustrated in Figure 2.3.
Interactive Visual Data Analysis
Published in Christian Tominski, Heidrun Schumann, Interactive Visual Data Analysis, 2020
Christian Tominski, Heidrun Schumann
Mosaic plots are suited for subsets with only one element, for example (a1),(a2),…. For the first step, the display area is split along the horizontal axis into rectangular segments. The number of segments corresponds to the number of distinct values of a1’s domain and the size of the segments represents the frequency of the values in a1’s value range. For the second step, each of the rectangular segments is split along the vertical axis with respect to a2. The procedure of splitting along the horizontal and vertical axes continues for all attributes in the nesting order. At the end, the rectangular segments of a mosaic plot represent the value distribution of the data. Figure 3.27 shows an example with the distribution of surviving passengers of the Titanic with respect to class and sex.
Class Maps for Visualizing Classification Results
Published in Technometrics, 2022
Jakob Raymaekers, Peter J. Rousseeuw, Mia Hubert
After a classification is carried out, we can display the result in a stacked bar chart or a mosaic plot (Hartigan and Kleiner 1981; Friendly 1994). Figure 1 shows such a stacked mosaic plot, which graphically represents the confusion matrix. The classes are represented by colors reminiscent of their meaning. The given classes are on the horizontal axis, and the predicted labels are on the vertical axis. The area of each rectangle is proportional to the number of objects in it. The display immediately shows that the given classes have different numbers of objects. Several variations of this plot are possible. One could rank the vertical labels in the order of the original classes, but we choose to put the given class at the bottom so that the lower part of each bar reflects the objects that were classified in accordance with their label. For the other labels in each bar we take the order of the remaining original classes. Here, we see that buds are often classified correctly but that there is some confusion between branch and support.