[Author Name(s), First M. Last, Omit Titles and Degrees]
[Include any grant/funding information and a complete correspondence address.]
Visual representation of data and understanding the methods to represent information such as data, conceptions, procedures, and associations in a graphical way is referred as Data visualization. Enormous data is generated globally on a daily basis, and Data visualization helps analyze such data and demonstrate it through info-graphics, which would be difficult using traditional methods. Data efficiency is achieved through low-dimensional space. Multidimensional scaling (MDS) and Principal Component Analysis (PCA) are the most significant methods used to decrease the dimensions. In the real world, most of the data is provided in raw format that is collected as asymmetric proximity matrix.
Any data that is captured in asymmetric matrix is converted to symmetric matrix, and the methods such as MDS or PCA are applied to it. Every asymmetric matrix has two directions; learning and reading that can be understood through Dynamic learning method, in which MDS is applied to the symmetrized conditions to get the primary representation. The dynamic learning method produces two representations maps; importing map (column vectors), and exporting map (row vectors). Information that cannot be represented without the use of virtualization can be easily represented by adding visual analysis tools such as cluster analysis and distance analysis, to these maps to know the data structure, and asymmetric data properties.
Dynamic learning method was presented to signify asymmetric data on metric spaces. Asymmetric relationships cannot be signified in an exact metric space, including 2D space. In metric space, data is symbolized as points, and generic links as symmetric. It was realized that MDS and PCA are not proficient in handling the asymmetric matrices, and the asymmetric properties of 2D space cannot be achieved by these methods. All these researches resulted in formation of dynamic learning methods that is useful to know the asymmetric proximity data.