Big data or bigdata can fall into various dimensions. k-means algorithms can provide results that are previously observed. It can also provide certain patterns on available data that has occurred over disparate units. However, to attain actual analysis of large volume of data and attain valuable information, a preset of data units charted on a graph and looking at its movement over, for instance TIME, cannot provide desired results or “validation”. Some times, ‘external” and “internal” evaluations could be looked at as a dialectical approach to “problem solving”. However, it may not fit within the realms of huge volume of data sets; big data ; bigdata. Further more, the clustering quality may depreciate. Above all that, the distance between centroids could cause issues in analysis. This has distance problem has already been foreseen and new set of algorithms has started emerging. BIRCH approach could result in new algorithms being put to test.
Approach II [In the process of writing]