Curse of dimensionality Data space becomes very sparse in high dimensions Volume of a hyper-sphere with largest range completely inside data space (i.e., a range of 0:5) is This is the probability that there is atleast one point within this hyper-sphere Hence, database should contain points