Application of Wavelet Analysis for Data Validation in Risk Assessment Systems
Keywords:
wavelet analysis, data validation, risk assessment, anomaly detection, component energy, correlation analysis, waveletAbstract
The object of the study is two information systems. The first system (S1) is a functioning system, the state of which can be normal or abnormal, described by a set of system parameters at a certain point in time. That is, there are time series describing the normal behavior of this system and deviations from it. The second system (S2) is designed to detect abnormal activity in the first system. The outputs of this anomaly detection system are presented by a time series, each of the values of which characterizes the state of the system at a certain point in time by two parameters — the confidence of the second anomaly detection system in assessing the state of the first system and the severity of the anomaly that occurred in the first system.
The study examines in detail the methodology for applying wavelet analysis, which includes the decomposition of the signal into approximating and detailing coefficients, signal reconstruction, calculation of the correlation between confidence and the severity system, as well as the analysis of the energy of the details to assess the instability of the system.
The proposed approach allows to set threshold values for determining instability in the operation of the anomaly detection system and develop metrics for the quality of its operation. Particular attention is paid to assessing the stability of the system's confidence through the energy of details, which characterizes the intensity of high-frequency changes in the signal, and determining the degree of consistency between confidence and severity through correlation analysis.
It is shown how this method integrates with the ACRAM risk assessment methodology, providing preliminary data validation, their filtering, comprehensive assessment and validation of results.
Based on the test data, the results of applying the developed algorithm are demonstrated, including the calculation of the energy of details, the correlation between confidence and severity, as well as setting thresholds to determine the level of system stability. The results obtained show that the anomaly detection system demonstrates an average level of stability, and correlation analysis reveals a significant negative correlation between the confidence of the system and the severity of the detected anomalies, which indicates a decrease in the confidence of the system during the transition to more serious states.
The proposed data validation method is an effective tool for increasing the reliability of risk assessment systems and can be used in various industries that require time series processing and anomaly detection.
References
V. Lytvyn, et al., “Fuzzy logic-based methodology for building access control systems based on fuzzy logic,” in 6th International Workshop on Modern Data Science Technologies, May, 31 — June, 1, 2024. [Electronic resource]. Available: https://ceur-ws.org/Vol-3723/paper7.pdf .
D.-W. Kim, G.-Y. Shin, and M.-M. Han, “Anomaly Detection Based on Discrete Wavelet Transformation for Insider Threat Classification,” Computer Systems Science and Engineering, no. 46 (1), pp. 153-164, 2023. https://doi.org/10.32604/csse.2023.034589 .
H. G. Mohamed, et al., “Optimal Wavelet Neural Network-Based Intrusion Detection in Internet of Things Environment,” no. 75 (2), pp. 4467-4483, 2023. https://doi.org/10.32604/cmc.2023.036822 .
M. Wan, Y. Song, Y. Jing, and J. Wang, “Function-Aware Anomaly Detection Based on Wavelet Neural Network for Industrial Control Communication,” Security and Communication Networks, Hindawi Limited, 5103270, 2018. https://doi.org/10.1155/2018/5103270 .
R. Purohit, S. Kumar, S. Sayyad, and K. Kotecha, “Time-frequency analysis and autoencoder approach for network traffic anomaly detection,” MethodsXM, vol. 14, 103228, 2025. https://doi.org/10.1016/j.mex.2025.103228 .
T. Rajesh, “Image Reconstruction Using Wavelet Transform with Extended Fractional Fourier Transform,” Blekinge Institute of Technology, 2014. [Electronic resource]. Available: https://www.diva-portal.org/smash/get/diva2:829956/FULLTEXT01.pdf .
Н. К. Смоленцев, Основи теорії вейвлетів. Вейвлети в MATLAB, вид 4-те, с. 102, 2014. [Електронний ресурс]. Режим доступу: https://surl.li/wnwoik .
Ортогональність. [Електронний ресурс]. Режим доступу: https://uk.wikipedia.org/wiki/%D0%9E%D1%80%D1%82%D0%BE%D0%B3%D0%BE%D0%BD%D0%B0%D0%BB%D1%8C%D0%BD%D1%96%D1%81%D1%82%D1%8C . Дата звернення 18.03.2025.
I. Daubechies, Ten Lectures of Wavelets, рp. 167-204, 1992. [Electronic resource]. Available: https://jqichina.wordpress.com/wp-content/uploads/2012/02/ten-lectures-of-waveletsefbc88e5b08fe6b3a2e58d81e8aeb2efbc891.pdf .
A. V. Oppenheim, R. W. Schafer, and J. R. Buck, Discrete-Time Signal Processing. Second edition, p. 41, 1998. [Electronic resource]. Available: https://ru.scribd.com/document/509418212/Discrete-Time-Digital-Signal-Processing-Oppenheim-Schafer-Buck .
O. Rioul, and P. Duhamel, “Fast Algorithms for Discrete and Continuous Wavelet Transforms,” IEEE Transactions on information theory, vol. 38, no. 2, p. 4, 1992. https://doi.org/10.1109/18.119724 .
Energy of a signal. [Electronic resource]. Available: https://www.gaussianwaves.com/2013/12/power-and-energy-of-a-signal/ . Accessed: 18.03.2025.
[[1]13] Стандартне відхилення. [Електронний ресурс]. Режим доступу: https://berg.com.ua/indicators-overlays/stdev . Дата звернення 18.03.2025.
Кореляція. [Електронний ресурс]. Режим доступу: https://uk.wikipedia.org/wiki/%D0%9A%D0%BE%D1%80%D0%B5%D0%BB%D1%8F%D1%86%D1%96%D1%8F . Дата звернення 18.03.2025.
Published
How to Cite
Issue
Section
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:
- Authors retain copyright and grant the journal right of first publication.
- 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 acknowledgment of its initial publication in this journal.
- 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 (See The Effect of Open Access).
