The evolution of individual and global learning preferences is influenced by correlation factors. This study introduces a novel evolutionary modeling approach to observe and analyze factors that affect the evolution of learning preferences. The influencing factors considered in this study are closely interwoven with the underlying personality of the students, individual traits, learning partners and interactions. This paper proposed non-mutually exclusive 3-state cellular automata evolution model that improves on previous approaches to study the evolution of learned preferences by overcoming the limitations of data acquisition through self-reported measurements or behavioral observations in a controlled environment. The experimental data is a large sample generated from the initial seed by the synthetic minority over-sampling technique (SMOTE) method. The seeds were derived from survey responses provided by 38 participants. The results demonstrate the varying degrees of influence of factors such as membership ratio, group size, membership distribution, and learning environment on the process and outcome of group preference evolution. The findings provide valuable insights into understanding how learning preferences evolve and how educators adapt to the learning environment. Furthermore, educators meeting the diverse learning preferences of students, that is, education that adapts to the dynamic demands of students, echoes the current educational trends of personalization and AI-driven learning.