Neural networks application for cluster analysis of the healthcare system crisis Cover Image

Застосування нейронних мереж для кластерного аналізу кризового стану системи охорони здоров’я
Neural networks application for cluster analysis of the healthcare system crisis

Author(s): Iryna Markina, Mohammed A. Y. Alshrafi
Subject(s): Economy, Welfare systems, Health and medicine and law, Family and social welfare
Published by: Institute of Society Transformation
Keywords: Healthcare System; Crisis; Cluster Analysis; Kohonen Self-organising Maps; Neural Networks; Regions;

Summary/Abstract: Introduction. Assessment of the status and trends of the healthcare system is a prerequisite for its effective management, control over the activities of institutions in healthcare, as well as development of effective measures to preserve and strengthen health of the population. Purpose of the study. To develop an approach for the healthcare system assessment on the basis of indicators that will help to determine its efficiency in modern conditions. Methods: logical, monographic, cluster and comparative analysis using Kohonen self-organising maps within Deductor Studio Academic software. Results. In the article, the approach to the analysis of the healthcare system by regions of Ukraine on the basis of Kohonen self-organising maps has been worked out. The algorithm of self-organising maps formation and stages of assessment, as well as interpretation of the results have been presented. The model formed is able to adapt quickly to new data inputs and does not require the involvement of experts to identify hidden patterns, and graphically present the results in user-friendly form. As the result of undertaken analysis, the clusters characterising the state of the Ukrainian healthcare system were worked out. It has been found that Kyiv, Volyn, Kirovograd and Chernihiv regions of Ukraine belong to the best cluster, whereas Chernivtsi and Khmelnytskyi regions - to the worst one. Conclusion and discussion. The study author proposed and applied the approach of the healthcare system assessment based on neural networks. In the long term, it may be possible to calculate the average value of each indicator to assess the crisis situation at the best and the worst cluster to develop recommendations on anti-crisis management activities to improve the situation. Further implementation of the proposed approach to the assessment of the healthcare system will help to select priorities for reform in the regions of Ukraine and carry out continuous monitoring of the healthcare system in the region.

  • Issue Year: 162/2016
  • Issue No: 11-12
  • Page Range: 56-61
  • Page Count: 6
  • Language: Ukrainian