Adaptive fast fuzzy clustering system

摘要: A parallel processing computer system for clustering data points in continuous feature space by adaptively separating classes of patterns. The preferred embodiment this massively includes preferably one processor per and requires a single priori assumption central tendency the distributions defining pattern classes. It advantageously exploits presence noise inherent gathering to not only classify into clusters, but also measure certainty classification each point, thereby identifying outliers spurious points. taught present invention is based upon gaps between successive values within features. This discrimination aspect achieved applying minimax comparison involving gap lengths locations largest smallest gaps. Clustering may be performed near-real-time on huge spaces having unlimited numbers