Cluster analysis is a statistical method for classifying samples or indicators. Which is an exploratory data analysis method. Cluster analysis groups and categorizes seemingly disordered objects (such as tables. People. Trees. Emotions. Concepts. Etc.). And classifies them according to the characteristics of individuals or samples. So that individuals in the same category have the highest possible probability. Homogeneity. And as high as possible heterogeneity between different categories/groups to better understand the research subject.
Birds of a feather flock together
With the help of cluster analysis algorithm. It can help Cyprus Phone Number us to spy on the data differences between different populations (figure 1). Therefore. This method is also applied in the practice of user classification based on quantitative data. The application of cluster analysis in user classification portraits——based on the application ideas and cases of psychostatistics Figure 1: schematic diagram of cluster analysis in two dimensions (2 variables) However. Since the data used in cluster analysis is not clearly classified. The number of categories after cluster analysis is also unknown.
That is. We do not know
the approximate distribution of the samples used for cluster analysis. Nor do we know which categories the system will classify them into. And there may not be any information about the categories for reference in advance. Therefore. Cluster analysis is more like a method for establishing hypotheses. And other statistical methods are needed to test related hypotheses. In the process of generating user portraits. It is recommended to use cluster analysis as a way to explore the classification structure and provide data support. Rather than (and may not) rely entirely on cluster analysis to form end-user classification conclusions.