Description Complete the Jupyter Notebook (available in the resources section), and if you can complement the solution by applying other clustering algorithms different from K-MEANS, the better. The idea is to find a solution that gives us a clustering as similar as possible to the original labeling of the data.
Phases of the project:
Phase 1: Apply a scaling on the dataset so that all features become on the same scale and the algorithm does not prioritize variables with higher variance. Phase 2: Apply the K-MEANS algorithm with a K=3 determined by the previous steps of silhouette and elbow method, which are already developed. Phase 3: Apply the PCA algorithm to transform the 4 columns of features to a new dataset of 2 main features, so that in the following steps the clusters can be observed in a 2-dimensional graph.
Additional step: If possible, apply algorithms other than K-means. To cluster the elements you can apply hierarchical clustering, dbscan, or whatever algorithm you want to test, the objective is to evaluate what results are obtained with different algorithms.