Work

New data clustering heuristic algorithm

Year: 2015
Type: article
Abstract: Clustering is the data mining technique that is used to place or collect objects into groups in such a way that objects in the same group are more similar or related among themselves than to those in other groups. These groups, called clusters, resemble each other but differ from other groups in objects which those contain. In this article the method of data clustering on the example of random data with uniform distribution was proposed. This article is focused on clustering in data mining. Data mining represents solving the problems by clustering large data sets with different data types and properties. The main task of the research was investigating data clustering and finding out how many clusters the data set contains. In particular, we were interested in answering the question whether there is more than one cluster in this data set. New method includes the decision rule. Decision rule uses the following parameters: area of regions found by the density distribution of input data, the number and magnitude of local maxima (peaks) found in each region, the number of elements (of the total number of primary elements) that fall into each found region. Proposed clustering method differs from existing, that the input parameter is the only data set and the criterion for evaluating the correctness of this method, is an objective assessment of a person or group of people based on visual logical analysis. All manipulations with the data mentioned in this article were made by using the Matlab software.
Language: English
Open Access status: gold
Estimated APC paid: $242
OpenAlex ID: W1556533528
FWCI:
Cites: 20
Cited by:
Related to: 8