ABSTRACTS OF ARTICLES OF THE JOURNAL "INFORMATION TECHNOLOGIES".
No. 1. Vol. 26. 2020

DOI: 10.17587/it.26.56-61

A. I. Lepsky, alexlep97@gmail.com, Saint Petersburg State University, Saint Petersburg, Russian Federation

Comparative Analysis of Leukocyte Clustering Algorithms According to FS and SS Parameters in a Cytofluorimetric Blood Test

In laboratory studies in biology and medicine, obtaining formal rules for evaluating the numerical values of empirical data is of great practical importance. One of the unsolved problems in the cytometric blood test is the automatic typology of leukocytes. A possible approach, in this case, could be the use of cluster analysis methods. However, with the clustering of white blood cells, many unresolved issues remain. The article explores various clustering algorithms. When conducting numerical experiments, it was shown that hierarchical methods and the K-means method do not give positive results. The DBSCAN method is the most promising for further study. A program code written in Python was created to conduct numerical experiments.
Keywords: cluster analysis, flow cytometry, machine learning

P. 56–61

To the contents