Control Systems and Computers, N2, 2019, Article 4
https://doi.org/10.15407/usim.2019.02.031
Upr. sist. maš., 2019, Issue 2 (280), pp. 31-39.
UDC 004.89
O.V. CHORNA, Master student of the Applied Mathematics Faculty, National Technical University of Ukraine “Igor Sikorsky Kyiv Politechnic Institute”, 03056, Peremohy Ave 37, Kyiv, Ukraine, chorna.oksana.v@gmail.com
L.A. LYUSHENKO, PhD, Senior Lecturer, Computer Systems Software Department of the Applied Mathematics Faculty, National Technical University of Ukraine “Igor Sikorsky Kyiv Politechnic Institute”, 03056, Peremohy Ave 37, Kyiv, Ukraine,
yushenkol@gmail.com
N.A. RYBACHOK, PhD, Senior Lecturer, Computer Systems Software Department of the Applied Mathematics Faculty, National Technical University of Ukraine “Igor Sikorsky Kyiv Politechnic Institute”, 03056, Peremohy Ave 37, Kyiv, Ukraine,
rybachok@pzks.fpm.kpi.ua
MODIFIED METHOD OF AUTOMATION DECISION-MAKING MANAGEMENT PROCESS FOR THE PROJECT MANAGEMENT TEAM CREATION
Introduction. One of the most important factors in the success of software development projects is the effective management of the human resources. For these purposes automated decision support systems are commonly used that can be based on data mining methods.
Purpose. The purpose of this research is to develop a modified method of automation decision-making management process based on data mining.
Methods. Modification of the method consists in using the restriction on the structure of the graph and considering such an extension of the Naive Bayes classifier, where each node can additionally have no more than one parent among other input variables. Using this modification, better accuracy will be achieved in assessing the selection of each team member, taking into account a large number of input parameters that describe the professional skills and competencies of the workers.
Results. To analyze the effectiveness of the proposed method modification, appropriate software was implemented using C# language. There was also implemented Naive Bayes algorithm in Matlab for comparison. Testing was carried out using a variety of personal cases. The modified method of automation decision-making management process of choosing a project team showed better results for accuracy criteria than the standard MatLab method, as well as a higher performance.
Conclusion. The presented modified method of automation decision-making management process based on data mining can be used for an efficient selection of the project team.
Download full text! (In Ukrainian)
Keywords: Data mining, Bayes’ theorem, Bayes’ network, classification, probability, human resource management.
REFERENCES
- Shapiro, V.D., Ilin, N.I., 1996. Project Management. San-Peterburg, Gazprom, 610 p. (In Russian).
- Keen P., 1980. Decision support systems: a research perspective. Cambridge, Massachusetts: Center for Information Systems Research, Alfred P. Sloan School of Management.
https://doi.org/10.1016/B978-0-08-027321-1.50007-9 - John Wiley & Sons – Data Science and Big Data Analytics: Discovering, Analyzing, Visualizing and Presenting Data. 2014, pp. 12-19.
- Maron M.E., 1961. Automatic Indexing: An Experimental Inquiry.
https://doi.org/10.1145/321075.321084 - Onyschuk, R.M., Tereschenko, I.M., 2015. Algorithmdescriptionforconstructing a NaiveBayesclassifier. (In Ukrainian).
- Pearl Judea. Causality: Models, Reasoning, and Inference. Cambridge University Press. 2000. ISBN 978-0-521-77362-1
- Zgurovsky, M.Z., Bidyuk, P.I., Terentyev, O.M., Prosyankina-Zharova, T.I., 2015. Bayesian Networks in Decision Support Systems. K.: Edelweis, 300 p. (In Ukrainian).
Received 01.04.2019
