Figure 7 shows the basic working model of our Neural Data Analysis tool, NDA [8]. First, the data can be preprosessed. Then it can be used for training of a SOM or statistical mappings can be applied to reduce dimensionality. The SOM and statistical values can be visualized in multiple ways and thereafter decided, if a new kind of preprosessing or analysis is needed. This whole process is interactive and allows the user to switch from visualization to preprosessing or dimensionality reduction and back.
Figure 7: An iterative and interactive model.