Summary

In this chapter, we presented the World Happiness Report data, providing a description of the data's purpose, as well as describing the data's properties. In order to gain further insights into the data, we utilized cluster analysis, leveraging ensemble techniques. We used co-occurrence matrix linkage in order to combine the cluster assignments of different base clusters. We tested various setups, with different ensemble sizes and numbers of neighbors, in order to provide a k-NN ensemble. After identifying that a t-SNE decomposition with a K value of 10 and 20 base clusters can be utilized, we analyzed the cluster assignments. We found that countries reporting the same happiness levels can, in fact, have different profiles. The most unhappy countries were, on average, developing countries who have to overcome many problems, concerning both their economies, and...