explained_variance_ratio_) plt.figure() plt.scatter(x_pca[:, 0][:int(len(X1))], x_pca[:, 1][:int(len(X1))], color='red', marker='^', alpha=0.5) plt.scatter(x_pca[:, 0][int(len(X2)):], x_pca[:, 1][int(len(X2)):], color='blue', marker='o', alpha=0.5) ...
In the previous section I explained how clustering on UMAP components can be more beneficial than clustering on tSNE or PCA components. However, if we decide to cluster on UMAP components, we need to be sure that distances between data points within clusters (local structure) and between cluster...