This paper uses the Apriori algorithm to mine association rules from user-based anime rating transactions. Support, confidence, and lift metrics are used to evaluate co-preference structures and identify strong item dependencies.

This project applies PCA and MDS to explore the multivariate structure of coffee sensory evaluation data. The analysis investigates variance structure, distance geometry, and species patterns in reduced dimensional space.

This project uses unsupervised learning to group ATP players from the 2023 season based on season-level performance data. By analyzing win rates, surface performance, match volume, and opponent ranking, players are segmented using clustering methods. The results show that tennis performance involves multiple dimensions and cannot be explained by ranking alone.