Modifications of the method for predicting sharp activity surges in systems with network effects using adaptive parameters

Cover Page

Cite item

Full Text

Open Access Open Access
Restricted Access Access granted
Restricted Access Subscription or Fee Access

Abstract

The article proposes modifications to the method for predicting abrupt changes in activity in software products with network effects, aimed at improving the accuracy and timeliness of detecting critical events. The focus is on adaptive parameters of the method, including a dynamic activation threshold for predictive signals dependent on current data volatility and automatic determination of the moving average window width based on local variability of activity metrics. А comparative analysis of the effectiveness of each modification was conducted using performance metrics defined in the article and visualization of results. It was found that the adaptive threshold reduces the proportion of false positives, while the algorithm for automatically selecting the moving average window width enables earlier detection of predictive signals. The results demonstrate that combining the proposed modifications ensures a balance between sensitivity and reliability of predictions, which is particularly important for social network monitoring systems and forecasting the risk of coordinated destructive actions by users.

Full Text

Restricted Access

About the authors

V. V. Ryabov

Federal state financed educational institution of higher education "Tambov State Technical University"

Author for correspondence.
Email: ryabov.vv@inbox.ru

PhD student

Russian Federation, Tambov

V. A. Nemtinov

Federal state financed educational institution of higher education "Tambov State Technical University"

Email: nemtinov.va@yandex.ru

Dr. of Tech. Sc., Professor

Russian Federation, Tambov

V. V. Alekseev

Federal state financed educational institution of higher education "Tambov State Technical University"

Email: vvalex1961@mail.ru

Dr. of Tech. Sc., Professor

Russian Federation, Tambov

References

  1. Wang H., Qiu L., Tan K., Cui J. Susceptible-Infectious-Critical-Recovered rumor spreading model in social networks, International Journal of Modern Physics C, 2022, vol. 33, no. 04, doi: 10.1142/s012918312250053x.
  2. Du R., Yu Z., Mei T., Wang Z. et al. Predicting activity attendance in event-based social networks: Content, context and social influence, Proceedings of the 2014 ACM international joint conference on pervasive and ubiquitous computing, 2014, pp. 425—434.
  3. Mancini A., Desiderio A., Di Clemente R., Cimini G. Self-induced consensus of Reddit users to characterise the GameStop short squeeze, Scientific Reports, 2022, vol. 12, no. 1, pp. 13780, doi: 10.1038/s41598-022-17925-2.
  4. Su Y., Cui C., Qu Н. Self-Attentive Moving Average for Time Series Prediction, Applied Sciences (Switzerland), 2022, vol. 12, no. 7, pp. 3602, doi: 10.3390/app12073602.
  5. Kuznetsov I. V., Rodkin M. V., Serebryakov D. V. Forecasting surges in serious crimes based on the hierarchy of crime patterns, Preprints of the Keldysh Institute of Applied Mathematics, 2005, no. 12, pp. 1—24.
  6. Certificate of State Registration of Computer Program No. 2025615409, Russian Federation. Search for predictive signals of a sharp increase in activity in a social network community based on comment data: No. 2025612926: applied 19.02.2025: published 04.03.2025 / V. V. Ryabov, V. A. Nemtinov; applicant: Tambov State Technical University (in Russian).
  7. Baumgartner J., Zannettou S., Keegan B. et al. The Pushshift Reddit Dataset, Proceedings of the International AAAI Conference on Web and Social Media, 2020, vol. 14, pp. 830—839, doi: 10.1609/icwsm.v14i1.7347.
  8. Anand A., Pathak J. The role of Reddit in the GameStop short squeeze, Economics Letters, 2022, vol. 211, pp. 110249, doi: 10.1016/j.econlet.2021.110249.
  9. Koroteev M. V. Fundamentals of Machine Learning in Python, Moscow, Knorus, 2024, 432 p. (in Russian).
  10. Korobova M. A., Gubina D. I. Overview of Machine Learning Methods Used in Algorithmic Trading, Languages in Professional Communication, 2022, pp. 54—59.
  11. Ryabov V., Nemtinov V. et al. Prospects for Predicting Coordinated Actions by Large Groups of Users Based on the Hierarchy of Activity in Social Networks, International Conference on Business Intelligence and Data Analytics (Singapore 2024), Singapore: Springer, 2024, pp. 15—26.
  12. Wagle А. Deep Learning for Financial Time Series using Long Short-Term Memory Model, International Journal of Science and Research, 2024, vol. 13, no. 4, pp. 1944—1972, doi: 10.21275/sr24418141736.
  13. Jammalamadaka S. R. Introduction to Linear Regression Analysis, American Statistician, 2003, vol. 57, no. 1, pp. 67.
  14. Gustavsen В. Passivity Enforcement by Residue Perturbation via Constrained Non-Negative Least Squares, IEEE Transactions on Power Delivery, 2021, vol. 36, no. 5, pp. 2758—2767, doi: 10.1109/tpwrd.2020.3026385.
  15. Wang J. Y., Yin Y. H., Zheng J. Yi. et al. Least absolute shrinkage and selection operator-based prediction of collision cross section values for ion mobility mass spectrometric analysis of lipids, The Analyst, 2022, vol. 147, no. 6, pp. 1236—1244, doi: 10.1039/d1an02161c.
  16. Ahmed H. A., Muhammad Ali P. J., Faeq A. K., Abdullah S. M. An Investigation on Disparity Responds of Machine Learning Algorithms to Data Normalization Method, ARO: The Scientific Journal of Koya University, 2022, vol. 10, no. 2, pp. 29—37, doi: 10.14500/aro.10970.
  17. Wieczorek J., Guerin C., McMahon T. K-fold cross-validation for complex sample surveys, Stat., 2022, vol. 11, no. 1, doi: 10.1002/sta4.454.
  18. Koleintev N. Ya., Goncharova O. A. Correlation and determination coefficients, Special Equipment and Transport Technologies, 2021, no. 10, pp. 206—212 (in Russian).
  19. Katrutsa A. M., Strizhov V. V. The problem of multicollinearity in feature selection for regression tasks, Information Technologies, 2015, vol. 21, no. 1, pp. 8—18 (in Russian).
  20. Zinzendoff Okwonu F., Laro Asaju B., Irimisose Arunaye F. Breakdown Analysis of Pearson Correlation Coefficient and Robust Correlation Methods, IOP Conference Series: Materials Science and Engineering, 2020, vol. 917, no. 1, pp. 012065, doi: 10.1088/1757-899x/917/1/012065.
  21. Bu C., Zhang Z. Research on Overfitting Problem and Correction in Machine Learning, Journal of Physics: Conference Series, 2020, vol. 1693, no. 1, pp. 012100, doi: 10.1088/1742-6596/1693/1/012100.
  22. Kozyrev А. N. Network effects and digital platforms in economics and mathematical models, Digital Economy, 2021, no. 3 (15), pp. 5—33, doi: 10.34706/DE-2021-03-01 (in Russian).
  23. Atchade M. N. Adaptive Forecasting Methods: Implementation in Excel and R, St. Petersburg, Saint Petersburg State University of Economics, 2018, 101 p. (in Russian).
  24. Ryabov V. V., Nemtinov V. A. Analysis of seasonality and trends in virtual social network actor activity using time series forecasting tools, Virtual Modeling, Prototyping and Industrial Design: Proceedings of the X International Conference (Tambov, 21—22 October 2024), Tambov, TSTU, 2024, pp. 280—286 (in Russian).
  25. Gonzalez-Torres M. A. Technology at the Rescue? Online Games, Adolescent Mental Health and the COVID Pandemic, The American Journal of Psychoanalysis, 2024, vol. 84, no. 2, pp. 190—202, doi: 10.1057/s11231-024-09445-x.
  26. Semenova V., Gorduza D., Wildi W. et al. Wisdom of the Crowds or Ignorance of the Masses? А Data-Driven Guide to WallStreetBets, Journal of Portfolio Management, 2024, vol. 50, no. 4, pp. 88—106, doi: 10.3905/jpm.2024.1.581.
  27. Liu Y., Li Y., Xie D. Implications of imbalanced datasets for empirical ROC-AUC estimation in binary classification tasks, Journal of Statistical Computation and Simulation, 2024, vol. 94, no. 1, pp. 183—203, doi: 10.1080/00949655.2023.2238235.

Supplementary files

Supplementary Files
Action
1. JATS XML
Download
2. Рис. 1. Forecast with a moving average window width equal to 3

Download (114KB)
Indexing metadata
3. Рис. 2. Forecast with a moving average window width equal to 30

Download (100KB)
Indexing metadata
4. Рис. 3. Plot with an example of a dynamically calculated adaptive threshold value b*

Download (86KB)
Indexing metadata
5. Рис. 4. Variation of the adaptive moving average window width in the computational experiment

Download (71KB)
Indexing metadata
6. Рис. 5. Plot comparing the results of modifications of the original method for forecasting abrupt changes in the system state by ROC-AUC and by formula (1); the most preferable result is marked with the symbol “S”

Download (94KB)
Indexing metadata

Copyright (c) 2026 Informacionnye Tehnologii


СМИ зарегистрировано Федеральной службой по надзору в сфере связи, информационных технологий и массовых коммуникаций (Роскомнадзор).
Регистрационный номер и дата принятия решения о регистрации СМИ: серия ПИ № 77 - 15565 от 02 июня 2003 г.