GRAVITATIONAL ATTRACTOR ENGINE FOR ADAPTIVELY AUTOCLUSTERING N-DIMENSIONAL DATA STREAMS

摘要

This invention relates to a method for autoclustering N-dimensional data streams, described as a gravitational attractor engine consisting of a gravitational attractor having a geometric boundary. The data stream (5) is analyzed for purposes of precisely centering each attractor's membership boundary surface (2) about the statistical center-of-mass (1) the data cluster it is intending to classify. The invention has particular utility in analyzing multi-parameter data from a flow cytometer, and more particularly has utility in analyzing data from whole blood cells tagged with fluorescently labelled CD3, CD4 and CD8 monoclonal antibodies to which a known number of fluorescent microbeads has been added.