Investigation of the capabilities of neural networks for predicting the functioning of message brokers of technological platforms

Igor Chirkov, Maxim Dunaev

Abstract


The problem of software failures in the operation of complex corporate software systems is economically significant and, unfortunately, inevitable. Therefore, to solve this problem, it is necessary to predict failures in a timely manner, based on information from the event logs (logs) of individual applications.

This paper investigates the capabilities of neural networks for predicting the performance of applications running on technological platforms and detecting anomalous situations. Network training is carried out on the basis of information from the log metrics of individual applications of technological platforms.

Were considered linear, dense, convolutional, recurrent neural networks, as well as the Holt-Winters and XGBoost models. The quality metric MAE (mean absolute error) is used to compare the results. As a result of comparisons, the best result was shown by recurrent neural networks (AR LSTM).

Full Text:

PDF (Russian)

References


Chandola V., Banerjee A., Kumar V. Anomaly Detection:A Survey // ACM Computing Surveys 41, Vol. 3, No. 15, 2009. pp. 1–58.

Rysin N. Development of a highly loaded time series forecasting solution: collection of scientific papers «Chronoeconomics » № 4(25). - Moscow 2020 – pp. 82-86.

Devyatkin A., Izvin A., Zykov D. Application of the method of group accounting of arguments in time series forecasting // collection of scientific papers «E-Scio» № 4(25). - Moscow 2020.

Indrajeet Y. Javeri, Mohammadhossein Toutiaee, Ismailcem B. Arpinar, and John A. Miller, Tom W. Miller Improving Neural Networks for Time-Series Forecasting using Data Augmentation and AutoML - University of Georgia, Athens, Georgia 2021.

Andryukhin E., Ridley M., Pravikov D. Prediction of failures and failures in distributed control systems based on time series forecasting models // Cybersecurity issues. – 2019. – № 3 (31). – pp. 24-32

Razvan-Gabriel Cirstea, Chenjuan Guo, Bin Yang Graph Attention Recurrent Neural Networks for Correlated Time Series Forecasting -- Full version - Aalborg 2021.

Selim Furkan Tekin, Oguzhan Karaahmetoglu, Fatih Ilhan, Ismail Balaban, Suleyman Serdar Kozat Spatio-temporal Weather Forecasting and Attention Mechanism on Convolutional LSTMs, Cornell University, 2021.

Alzheev A., Kochkarov R. Comparative analysis of predictive models ARIMA and LSTM on the example of Russian stocks. Finance: theory and practice. 2020;24(1):14-23. DOI: 10.26794/2587-5671-2020-24-1-14-23

Blednov R., Skvortsov N. Predicting application crashes based on the state logs of software objects // Modern information technology and IT education. 2019. Т. 15, № 4. С. 945-953. DOI: 10.25559/SITITO.15.201904.945-953

Nikolenko S., Kadurin A., Arkhangelskya E. Deep learning. — St. Petersburg.: Peter, 2018. — 480 p.: illustrated. — (Series « Programmer's library »).

Barsky A. Logical neural networks / A. Barsky. - M.: Binomial. Knowledge laboratory / Internet University of Information Technologies (INTUIT), 2017. - 336 p.

Galushkin A. Neural networks: basic theory / A. Galushkin. - M.: editorial and publishing system, 2015. - 496 p.

Khaikin S. Neural Networks: Complete Course / S. Khaikin. - M.: Dialectics, 2019. - 1104 p.

Maylawati, Dian S., Saptawati P. Set of Frequent Word Item sets as Feature Representation for Text with Indonesian Slang // Journal of Physics: Conference Series, Vol. 801, No. 1, 2017. P. 012066.

Andreas, Muller An introduction to machine learning with Python. Data Scientist's Guide // monograph. / Muller Andreas. - M.: Alpha-book, 2017


Refbacks

  • There are currently no refbacks.


Abava  Кибербезопасность MoNeTec 2024

ISSN: 2307-8162