SOME APPROACHES TO THE ANALYSIS OF DATA IN ARCHEOLOGY
Yu.P. Kholyushkin1, V.S. Kostin2
1State Public Scientific Technological Library of the SB RAS, Voshod str., 15, Novosibirsk, 630200, Russia
2Institute of Economics and Industrial Engineering of the SB RAS, Academician Lavrentyev Ave., 17, Novosibirsk, 630090, Russia
Keywords: дисперсионный анализ, кластерный анализ, метод повторной выборки с возвращением, сравнение классификаций, обобщенная классификация, analysis of variance, cluster analysis, a method of re-sampling with replacement, comparing of classifications, generalized classification
Abstract
In order to accomplish up-to-date tasks, archeology must use one of the recent areas of Artificial Intelligence - «intellectual data analysis» which is a brief and imprecise translation from English of the terms «Data Mining» and «Knowledge Discovery in Databases» (DM & KDD). For several years, the authors have been conducting research on the statistical analysis of data in archeology. In the course of these studies we have developed a coherent application of methods in archeology with the analysis of citation and identification of scientific schools. In the course of the studies we used: the methods of analysis of variance; methods of factor analysis and multidimensional scaling; clustering. With the help of clustering, Data Mining allocates various homogeneous groups of data; a procedure to detect the structure of the table. The article provides specially designed tools for ordering an inhomogeneous archaeological information and identifying its structure. The method of re-sampling with replacement. Comparison of the classifications and forming of a generalized classification. In addition, the article works out the method of forming a generalized consolidated classification, based on an analysis of coincidence of different classifications of the same object.
|
