CLASSIFICATION OF SAMPLES DATA

摘要

FIELD: physics.SUBSTANCE: invention relates to classification of biomolecular data. For this, classification system (100) is used. System input receives multiple attributes (102) of the sample to be classified and multiple appropriate error estimates (104). Statistical module (106) associates probability distribution density functions (108) with said attributes. Note here that appropriate probability distribution density functions depend upon error estimates. Replication module (110) generates multiple disturbed dummy treatments (112). Note here that attributes are arbitrarily disturbed in compliance with apt probability distribution density functions. Classifier (114) classifies disturbed dummy treatments on the basis of disturbed attributes. Analyser (118) classifies the sample to be classified proceeding from the analysis of classified dummy treatments (116) to produce classification (120) of samples.EFFECT: higher reliability.13 cl, 6 dwg