Investigation of the performance of neural gas networks in hydrological clustering

Document Type : Research Paper

Authors

1 M.Sc., Department of Water Engineering, College of Agriculture, Isfahan University of Technology, Isfahan, Iran.

2 Professor, Department of Water Engineering, College of Agriculture, Isfahan University of Technology, Isfahan, Iran.

3 Assistant Professor, Department of Water Engineering, College of Agriculture, Isfahan University of Technology, Isfahan, Iran.

Abstract

The design of many infrastructures and construction projects requires the extensive studies in the area's geographical conditions and climatic characteristics. The effectiveness of this research itself depends on the information and data required. In many cases, the project area is in a situation where no climatic information such as rainfall is available. Hence, regional frequency analysis has been received much attention. In this way, having specific conditions and mechanisms, the available information in other sites can be expanded and transferred to the other areas. In this research, clustering is one of the most effective steps that divide the existing  stations into the hydrologically  homogeneous areas. Therefore, in this study, in addition to the common methods in clustering, two new models, neural gas network and growing neural gas network, were used to determine the homogeneous regions in Khuzestan province, Iran. One of the unique features of these algorithms is learning the topology or shape of the distributions that governs the data space. Using the variables of longitude, latitude, altitude, mean annual rainfall, and maximum 24-hour rainfall of the station, the design area was divided into two hydrologically homogeneous areas, and the clustering process was performed. The results show that the neural gas networks has a high efficiency and accuracy in view of clustering. The Mean Percentage Difference and Coefficient of Variation of Root Mean Square Error in neural gas were estimated to be 15.56 and 24.39 percent, respectively, which showed a considerable advantages over the conventional methods.

Keywords

Main Subjects


  1. Abdi, A., Hassanzadeh, Y., & Ouarda, T. B. (2017). Regional frequency analysis using Growing Neural Gas network. Journal of Hydrology550, 92-102.
  2. Adib, A., Kashani, A., & Ashrafi, S. M. (2021). Merge L-Moment Method, Regional Frequency Analysis and SDI for Monitoring and Zoning Map of Short-Term and Long-Term Hydrologic Droughts in the Khuzestan Province of Iran. Iranian Journal of Science and Technology, Transactions of Civil Engineering45(4), 2681-2694.
  3. Alemaw, B. F., & Chaoka, R. T. (2016). Regionalization of Rainfall Intensity-Duration-Frequency (IDF) Curves in Botswana. Journal of Water Resource and Protection, 8(12), 1128.
  4. Angelopoulou, A., Psarrou, A., Garcia-Rodriguez, J., Orts-Escolano, S., Azorin-Lopez, J., & Revett, K. (2015). 3D reconstruction of medical images from slices automatically landmarked with growing neural models. Neurocomputing150, 16-25.
  5. Ariff, N. M., Jemain, A. A., & Bakar, M. A. A. (2016). Regionalization of IDF curves with L-moments for storm events. International Journal of Mathematical and Computational Sciences10(5), 217-223.
  6. Caliński, T., & Harabasz, J. (1974). A dendrite method for cluster analysis. Communications in Statistics-theory and Methods3(1), 1-27.
  7. Carlevarino, A., Martinotti, R., Metta, G., & Sandini, G. (2000, July). An incremental growing neural network and its application to robot control. In Proceedings of the IEEE-INNS-ENNS International Joint Conference on Neural Networks. IJCNN 2000. Neural Computing: New Challenges and Perspectives for the New Millennium, 5, 323-328). IEEE.
  8. Chaubey, I., Haan, C. T., Grunwald, S., & Salisbury, J. M. (1999). Uncertainty in the model parameters due to spatial variability of rainfall. Journal of Hydrology220(1-2), 48-61.
  9. Chou, C. H., Su, M. C., & Lai, E. (2004). A new cluster validity measure and its application to image compression. Pattern Analysis and Applications7(2), 205-220.
  10. Cselényi, Z. (2005). Mapping the dimensionality, density and topology of data: The growing adaptive neural gas. computer methods and programs in biomedicine78(2), 141-156.
  11. de Oliveira Martins, L., Silva, A. C., De Paiva, A. C., & Gattass, M. (2009). Detection of breast masses in mammogram images using growing neural gas algorithm and Ripley’s K function. Journal of Signal Processing Systems,55(1),77-90.
  12. Decker, R. (2005). Market basket analysis by means of a growing neural network. The International Review of Retail, Distribution and Consumer Research15(2), 151-169.
  13. Durrans, S. R., & Kirby, J. T. (2004). Regionalization of extreme precipitation estimates for the Alabama rainfall atlas. Journal of Hydrology295(1-4), 101-107.
  14. Ferrer, G. J. (2014). Creating Visual Reactive Robot Behaviors Using Growing Neural Gas. In MAICS(pp. 39-44).
  15. Fink, O., Zio, E., & Weidmann, U. (2015). Novelty detection by multivariate kernel density estimation and growing neural gas algorithm. Mechanical Systems and Signal Processing50, 427-436.
  16. Fišer, D., Faigl, J., & Kulich, M. (2013). Growing neural gas efficiently. Neurocomputing,104, 72-82.
  17. Fritzke, B. (1995). A growing neural gas network learns topologies. Advances in neural information processing systems7, 625-632.
  18. Ghadami, M., Raziei, T., Amini, M., & Modarres, R. (2020). Regionalization of drought severity–duration index across Iran. Natural Hazards, 103(3), 2813-2827.
  19. Goovaerts, P. (1999). Using elevation to aid the geostatistical mapping of rainfall erosivity. Catena34(3-4), 227-242.
  20. Hosking, J.R.M., & Wallis, J.R. (1993). Some statistics useful in regional frequency analysis. Water resources research29(2), 271-281.
  21. Lee, S. H., & Maeng, S. J. (2003). Frequency analysis of extreme rainfall using L‐moment. Irrigation and Drainage: The journal of the International Commission on Irrigation and Drainage52(3), 219-230.
  22. Linda, O., & Manic, M. (2009, June). GNG-SVM framework-classifying large datasets with Support Vector Machines using Growing Neural Gas. In 2009 International Joint Conference on Neural Networks. (pp. 1820-1826). IEEE.
  23. Lisboa, P. J., Edisbury, B., & Vellido, A. (2000). Business applications of neural networks: the state-of-the-art of real-world applications(Vol. 13). World scientific.
  24. Martinetz, T., & Schulten, K. (1991). A" neural-gas" network learns topologies.
  25. Masselot, P., Chebana, F., & Ouarda, T. B. (2017). Fast and direct nonparametric procedures in the L-moment homogeneity test. Stochastic Environmental Research and Risk Assessment, 31(2), 509-522.
  26. Modarres, R. (2010). Regional dry spells frequency analysis by L-moment and multivariate analysis. Water resources management24(10), 2365-2380.
  27. Moreli, V., Cazorla, M., Orts-Escolano, S., & Garcia-Rodriguez, J. (2014, July). 3d maps representation using gng. In 2014 International Joint Conference on Neural Networks (IJCNN)(pp. 1482-1487). IEEE.
  28. Quintana-Pacheco, Y., Ruiz-Fernández, D., & Magrans-Rico, A. (2014). Growing Neural Gas approach for obtaining homogeneous maps by restricting the insertion of new nodes. Neural networks54, 95-102.
  29. Rousseeuw, P. J. (1987). Silhouettes: a graphical aid to the interpretation and validation of cluster analysis. Journal of computational and applied mathematics20, 53-65.
  30. Soltani, S., Helfi, R., Almasi, P., & Modarres, R. (2017). Regionalization of rainfall intensity-duration-frequency using a simple scaling model. Water Resources Management, 31(13), 4253-4273.
  31. Zaki, S. M., & Yin, H. (2008). A semi-supervised learning algorithm for growing neural gas in face recognition. Journal of Mathematical Modelling and Algorithms, 7(4), 425-435.