Анализ особенностей человеческого восприятия в полиметрических пространствах при оценивании графических профилей данных в системах экологического мониторинга

by

Yu.V. Doronina1, A.V. Skatkov2

1Sevastopol State University, RF, Sevastopol, Universitetskaya St., 33

2Institute of Natural and Technical Systems, RF, Sevastopol, Lenin St., 28

DOI: 10.33075/2220-5861-2026-1-129-138

UDC 519.8           

EDN: https://elibrary.ru/lcjbap

Abstract:    

The article presents a comprehensive study of methods for analyzing graphical profiles of monitoring objects in environmental control systems, taking into account the peculiarities of human perception based on polymetric space. The use of polymetric spaces makes it possible to integrate heterogeneous metrics into a single evaluation system, increasing the flexibility and completeness of the analysis. Statistical, frequency, and time approaches are considered, including classical and hybrid metrics such as Kolmogorov-Smirnov, Kulback-Leibler, and dynamic time alignment. A mathematical optimization model is proposed to approximate numerical estimates to subjective expert data. It is shown that using a hybrid metric with linear and nonlinear components can significantly improve the accuracy and interpretability of the results, especially in conditions of variability and noise of the data. The results are confirmed by visualization and quantification of correlations with human perception, which ensures objective verification of the proposed methods. The study demonstrates the prospects for practical application of the proposed approach in environmental control systems, biomedical analysis, computer vision and other areas where it is important to adapt automatic assessments to subjective criteria and take into account the human factor in the decision-making process in conditions of uncertainty of assessments.

Keywords: graphical profile, curve comparison metrics, human perception, hybrid metric, polymetric space

 REFERENCES

  1. Seleznev I.V. and Konopatskii E.V. Chislovye kriterii otsenki skhodstva mnogomernykh ge-ometricheskikh ob”ektov (Numerical criteria for similarity assessment of multidimensional geometric objects). *Trudy mezhdunarodnoi konferentsii «Grafikon-2022» (Proceedings of the International Con-ference “GraphiCon-2022”)*. 2022, pp. 85–92.
  2. Sokolova M. and Lapalme G. A systematic analysis of performance measures for classification tasks. Information Processing & Management, 2009, Vol. 45, No. 4, pp. 427–437.
    https://doi.org/10.1016/j.ipm.2009.03.002
  3. Fawcett T. An introduction to ROC analysis. Pattern Recognition Letters, 2006, Vol. 27, No. 8, pp. 861–874. https://doi.org/10.1016/j.patrec.2005.10.010
  4. Vedaldi A., et al. Metrics for graph comparison: A practitioner’s guide. PLOS One, 2020, Vol. 15, p. e0228728. https://doi.org/10.1371/journal.pone.0228728
  5. Bengio Y., Courville A., and Vincent P. Representation learning: a review and new perspec-tives. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2013, Vol. 35, No. 8, pp. 1798–1828. https://doi.org/10.1109/TPAMI.2013.50
  6. Wang C., et al. Unsupervised learning predicts human perception and memory through distributed representation. PLOS Computational Biology, 2021, Vol. 17, p. e1007934. https://doi.org/10.1371/journal.pcbi.1007934
  7. Storrs K.R., Anderson B.L., and Fleming R.W. Unsupervised learning predicts human perception and misperception of gloss. Nature Human Behaviour, 2021, Vol. 5, pp. 1402–1417.
    https://doi.org/10.1038/s41562-021-01097-6
  8. Li Y., Lou R., and Gao Y. Fast and accurate time series stream monitoring with dynamic time warp-ing. Proceedings of the VLDB Endowment, 2023, Vol. 16, No. 7, pp. 1605–1618.
    https://doi.org/10.14778/3594512.3594517
  9. Li K., et al. Using dynamic time warping self-organizing maps to identify typical diurnal patterns in time series environmental exposure data. Scientific Reports, 2021, Vol. 11, p. 24033.
    https://doi.org/10.1038/s41598-021-03515-1
  10. Du Y., et al. Hybrid approaches enhance hydrological model usability across diverse climate gra-dients. Nature Communications, 2025, Vol. 16, p. 2324. https://doi.org/10.1038/s43247-025-02324-y
  11. Moeini R., et al. Hybrid methods for reservoir operation rule curve determination under uncer-tainty. Journal of Hydrology, 2022, Vol. 610, p. 128996. https://doi.org/10.1016/j.jhydrol.2022.128996
  12. Ding H., et al. Case study in the Minjiang river Basin: ecosystem stability analysis using DTW of vegetation time series. Ecological Indicators, 2023, Vol. 154, p. 110451.
    https://doi.org/10.1016/j.ecolind.2023.110451
  13. Naeemah A.J., et al. Sustainability metrics and hybrid decision-making model for sustainable tool selection. Sustainable Futures, 2023, Vol. 5, p. 100266. https://doi.org/10.1016/j.sftr.2023.100266
  14. Guillaumot C., et al. Classic or hybrid? The performance of next generation ecological niche modeling techniques. Diversity and Distributions, 2022, Vol. 28, pp. 2684–2698.
    https://doi.org/10.1111/ddi.13617
  15. Bringmann K., et al. Dynamic Dynamic Time Warping. *Proceedings of the 2024 Annual ACM-SIAM Symposium on Discrete Algorithms (SODA)*, 2024, p. 10. https://doi.org/10.1137/1.9781611977912.10
  16. Koldasbayeva D., Tregubova P., Gasanov M., Zaytsev A., Petrovskaia A., and Burnaev E. Challenges in data-driven geospatial modeling for environmental research and practice. Nature Com-munications, 2024, Vol. 15, p. 5240. https://doi.org/10.1038/s41467-024-55240-8

Loading