METHOD FOR DATA CLASSIFICATION BASED ON MANIFOLD LEARNING

摘要

A method for classifying data based on manifold learning according to the present invention includes the following steps: removing an outlier from the whole collected data; building a variance function based on the data from which the outlier has been removed; projecting the whole data to a manifold space through a dynamic system based on the variance function; and applying a manifold learning algorithm based on the data obtained according to the projection step.