The subject of this repository was to perform basic cluster analysis on a seed dataset. The dataset contains geometrical properties of kernels belonging to three different varieties of wheat. We performed four different clustering approaches and obtained these results:
The agglomerative clustering turned out to be the best choice. DBSCAN had the lowest performance, however it looks more suitable for detecting outliers in datasets.