Abstract:

Today, thanks to microdis technology, it has become possible to express different levels of genes simultaneously. Representation of secret information within the genes makes it easier to understand the genes, but the excess of genes and the high volume of noise in data sets make it difficult to understand the genes. It is used to facilitate the understanding of the genes. Micro data is one of the best examples of multi-dimensional data. In the framework of the study to accumulate multi-dimensional data, new methods for the selection of starting accumulation centers for standard K-means and PSO accumulation algorithms have been proposed. It is also adapted to the PSO algorithm. The accuracy of the methods developed has been tested on the data sets that are commonly used in literature and these approaches have been implemented on the Colon Cancer microdis data sets. Based on the standard K-means and PSO accumulation methods, the approaches developed were compared; the performance was assessed using the average number of iterations achieved to solve, Rand and Silhouette index. Experimental studies have observed that developed approaches have given successful results on the normalized data sets that have been selected.

Keywords: