Clustering for Classification: Using Standard Clustering Methods to Summarise Datasets with Minimal Loss of Classification Accuracy