Data dimensionality reduction and evaluation of clusterization quality in the problems of analysis of composition of multi-component solutions.

Autor: Gushchin, K., Burikov, S., Dolenko, T., Persiantsev, I., Dolenko, S.
Zdroj: Optical Memory & Neural Networks; Jul2015, Vol. 24 Issue 3, p186-192, 7p
Abstrakt: This paper presents the results of search for optimal combination of a method of data dimensionality reduction and a clusterization algorithm, for analysis of an array of Raman spectra of multicomponent solutions of inorganic salts. The most informative criterion of evaluation of the quality of the obtained clusterization is presented. It is shown that application of special algorithms in combination with methods of dimensionality reduction improves the quality and increases the stability of solution of the clusterization problem. [ABSTRACT FROM AUTHOR]
Databáze: Complementary Index
Externí odkaz: