Study of Recognition Approach for Specific Sample Points in High Dimension Space

Abstract

Based on secondary analysis techniques to identify specific sample point using partial least-squares analysis method, the recognition method of specific sample point of two-dimensional floor plan of ellipse T <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> was extended to three-dimensional figure of ellipsoid T <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> and high-dimensional space of hyper- ellipsoid T <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> . Another Identification method of specific sample point making use of hierarchical diagram in high-dimensional space based on hierarchical clustering method was proposed at the same time. The recognition method of specific sample point had great significance on research areas of data mining, machine learning and pattern recognition while eliminating samples generated due to random factors and refining mathematical models. Empirical analysis of the five kinds of identification method was accomplished using ecological data of the 56 observation sites along the Bohai Sea coastal zone.