Modeling Financial Risk in Telecommunication Field

Authors

DOI:

https://doi.org/10.20535/1810-0546.2017.5.110338

Keywords:

Telecommunication company, Financial risks, Survival models, Gradient busting, Neural network, Logistic regression

Abstract

Background. The telecommunication field in Ukraine is dynamically developing continuously renewing its proposals for the market and consumer requirements. That is why a timely estimation of financial risks and optimization of financial expenses regarding development of new components and possible losses of clients is especially urgent problem today.

Objective. The aim of the paper is to suggest an approach for estimation of financial risks and forecasting of the client loss and optimal service time utilization based on intellectual data analysis and behavior models.

Methods. To determine the probability of customer loss the neural networks theory, gradient busting, random forest and logistic regression are used. The survival analysis models for possible client transition time to another company are developed.

Results. The best model for forecasting the clients intending for transition to another telecommunication company turned out to be the one based on gradient busting.

Conclusions. It was shown that timely estimation of financial losses, provoked by possible loss of clients, is an urgent task for intellectual data analysis. A perspective approach for optimization of the company financial resources is determining the time period related to possible loss of clients.

References

Google Determined the Way Ukrainians use Internet [Online]. Available: http://watcher.com.ua/2016/09/13/google-z-yasuvav-yak-ukrayintsi-korystuyutsya-internetom/ (in Ukrainian).

K. Vorontsov, Mathematical Learning Methods on Precedents. Course of Lectures. Moscow, Russia: MPTI, 2006 (in Russian).

P.I. Bidyuk et al., Analysis of Time Series. Kyiv, Ukraine: NTUU KPI, 2013 (in Ukrainian).

N.G. Zagoruyko, Applied Data and Knowledge Analysis Methods. Novosibirsk, Russia: Institute of Mathematics Publ., 1999 (in Russian).

I.A. Chubukova, Data Mining. Moscow, Russia: Binom LBZ, 2008.

S.O. Dovgij et al., Decision Making Systems on Probability-Statistic Methods. Kyiv, Ukraine: Logos, 2014 (in Ukrainian).

N.G. Zagoruyko et al., Algorithms of Empirical Patterns Detection. Novosibirsk, SU: Nauka, 1985 (in Russian).

D.R. Cox, “Regression models and life-tables”, J. Royal Statist. Soc. Ser. B (Methodological), vol. 34, no. 2, pp. 187–220, 2007.

O.V. Fomin and N.V. Kuznietsova, “Scoring models of credit card holder behavior for their solvency estimation”, Systemni Nauky ta Kibernetyka, no. 5, pp. 56–67, 2016 (in Ukrainian).

M. Marimo. (2005, Apr. 12). Survival Analysis of Bank Loans and Credit Risk Prognosis [Online]. Available: http://wiredspace.wits.ac.za/jspui/bitstream/10539/18597/1/Mercy%20Marimo%20Thesis_Survival%20Analysis_28.03.%202015_v1.pdf

N.V. Kuznietsova et al., “Using the Methodology of Survivability analysis for consumer risks researching”, Systemni Nauky ta Kibernetyka, no. 6, pp. 126–135, 2017 (in Ukrainian).

Downloads

Published

2017-10-31

Issue

No. 5 (2017): Engineering

Section

Art

License

Copyright (c) 2017 Igor Sikorsky Kyiv Polytechnic Institute

Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 International License.

Authors who publish with this journal agree to the following terms:

    1. Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under CC BY 4.0 that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.

    2. Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.

    3. Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work