Gaussian Representations of K-Means Clusters: Case Study of Educational Process Mining of UCI
The information of data patterns can help determining analytical direction, choosing right tools for analysis, and validating inferential results. However, this argument might not be helpful because of diverse patterns. To disclose inside information, a learning approach about data clustering is proposed by integrating K-means and Gaussian representation from data science. It gets insight of similar and dominant distribution through iterative learning. Its core technique lies in the design of data representation which can carry similarity and dominance characteristics from samples to K-learning. For illustration, it is applied in the educational process mining of UCI. Its results can provide strategic information for educational activities.