The Choice of an Appropriate Information Dissimilarity Measure for Hierarchical Clustering of River Streamflow Time Series, Based on Calculated Lyapunov Exponent and Kolmogorov Measures
Аутори
Mihailović, Dragutin T.Nikolić-Đorić, Emilija
Malinović-Milićević, Slavica
Singh, Vijay P.
Mihailović, Anja
Stošić, Tatjana
Stošić, Borko
Drešković, Nusret
Чланак у часопису (Објављена верзија)
Метаподаци
Приказ свих података о документуАпстракт
The purpose of this paper was to choose an appropriate information dissimilarity measure for hierarchical clustering of daily streamflow discharge data, from twelve gauging stations on the Brazos River in Texas (USA), for the period 1989–2016. For that purpose, we selected and compared the average-linkage clustering hierarchical algorithm based on the compression-based dissimilarity
measure (NCD), permutation distribution dissimilarity measure (PDDM), and Kolmogorov distance (KD). The algorithm was also compared with K-means clustering based on Kolmogorov complexity (KC), the highest value of Kolmogorov complexity spectrum (KCM), and the largest Lyapunov exponent (LLE). Using a dissimilarity matrix based on NCD, PDDM, and KD for daily streamflow,
the agglomerative average-linkage hierarchical algorithm was applied. The key findings of this study are that: (i) The KD clustering algorithm is the most suitable among others; (ii) ANOVA analysis shows that there exist highly significant d...ifferences between mean values of four clusters, confirming that the choice of the number of clusters was suitably done; and (iii) from the clustering we found that the predictability of streamflow data of the Brazos River given by the Lyapunov time (LT), corrected for randomness by Kolmogorov time (KT) in days, lies in the interval from two to five days.
Кључне речи:
streamflow time series / Brazos River / average-linkage clustering hierarchical algorithm / K-means clustering / Kolmogorov complexity-based measures / largest Lyapunov exponent / Lyapunov time / Kolmogorov time / predictability of streamflow time seriesИзвор:
Entropy, 2019, 21, 2, 215-Издавач:
- Switzerland, Basel : MDPI
Финансирање / пројекти:
- Истраживање климатских промена и њиховог утицаја на животну средину - праћење утицаја, адаптација и ублажавање (RS-MESTD-Integrated and Interdisciplinary Research (IIR or III)-43007)
Напомена:
- We acknowledge the support of Brazilian agencies CAPES and CNPq (grant no. 310441/2015-3).
DOI: 10.3390/e21020215
ISSN: 1099-4300
WoS: 000460742200116
Scopus: 2-s2.0-85061968979
Институција/група
Географски институт „Јован Цвијић“ САНУ / Geographical Institute Jovan Cvijić SASATY - JOUR AU - Mihailović, Dragutin T. AU - Nikolić-Đorić, Emilija AU - Malinović-Milićević, Slavica AU - Singh, Vijay P. AU - Mihailović, Anja AU - Stošić, Tatjana AU - Stošić, Borko AU - Drešković, Nusret PY - 2019 UR - https://dais.sanu.ac.rs/123456789/12914 AB - The purpose of this paper was to choose an appropriate information dissimilarity measure for hierarchical clustering of daily streamflow discharge data, from twelve gauging stations on the Brazos River in Texas (USA), for the period 1989–2016. For that purpose, we selected and compared the average-linkage clustering hierarchical algorithm based on the compression-based dissimilarity measure (NCD), permutation distribution dissimilarity measure (PDDM), and Kolmogorov distance (KD). The algorithm was also compared with K-means clustering based on Kolmogorov complexity (KC), the highest value of Kolmogorov complexity spectrum (KCM), and the largest Lyapunov exponent (LLE). Using a dissimilarity matrix based on NCD, PDDM, and KD for daily streamflow, the agglomerative average-linkage hierarchical algorithm was applied. The key findings of this study are that: (i) The KD clustering algorithm is the most suitable among others; (ii) ANOVA analysis shows that there exist highly significant differences between mean values of four clusters, confirming that the choice of the number of clusters was suitably done; and (iii) from the clustering we found that the predictability of streamflow data of the Brazos River given by the Lyapunov time (LT), corrected for randomness by Kolmogorov time (KT) in days, lies in the interval from two to five days. PB - Switzerland, Basel : MDPI T2 - Entropy T1 - The Choice of an Appropriate Information Dissimilarity Measure for Hierarchical Clustering of River Streamflow Time Series, Based on Calculated Lyapunov Exponent and Kolmogorov Measures SP - 215 VL - 21 IS - 2 DO - 10.3390/e21020215 UR - https://hdl.handle.net/21.15107/rcub_dais_12914 ER -
@article{ author = "Mihailović, Dragutin T. and Nikolić-Đorić, Emilija and Malinović-Milićević, Slavica and Singh, Vijay P. and Mihailović, Anja and Stošić, Tatjana and Stošić, Borko and Drešković, Nusret", year = "2019", abstract = "The purpose of this paper was to choose an appropriate information dissimilarity measure for hierarchical clustering of daily streamflow discharge data, from twelve gauging stations on the Brazos River in Texas (USA), for the period 1989–2016. For that purpose, we selected and compared the average-linkage clustering hierarchical algorithm based on the compression-based dissimilarity measure (NCD), permutation distribution dissimilarity measure (PDDM), and Kolmogorov distance (KD). The algorithm was also compared with K-means clustering based on Kolmogorov complexity (KC), the highest value of Kolmogorov complexity spectrum (KCM), and the largest Lyapunov exponent (LLE). Using a dissimilarity matrix based on NCD, PDDM, and KD for daily streamflow, the agglomerative average-linkage hierarchical algorithm was applied. The key findings of this study are that: (i) The KD clustering algorithm is the most suitable among others; (ii) ANOVA analysis shows that there exist highly significant differences between mean values of four clusters, confirming that the choice of the number of clusters was suitably done; and (iii) from the clustering we found that the predictability of streamflow data of the Brazos River given by the Lyapunov time (LT), corrected for randomness by Kolmogorov time (KT) in days, lies in the interval from two to five days.", publisher = "Switzerland, Basel : MDPI", journal = "Entropy", title = "The Choice of an Appropriate Information Dissimilarity Measure for Hierarchical Clustering of River Streamflow Time Series, Based on Calculated Lyapunov Exponent and Kolmogorov Measures", pages = "215", volume = "21", number = "2", doi = "10.3390/e21020215", url = "https://hdl.handle.net/21.15107/rcub_dais_12914" }
Mihailović, D. T., Nikolić-Đorić, E., Malinović-Milićević, S., Singh, V. P., Mihailović, A., Stošić, T., Stošić, B.,& Drešković, N.. (2019). The Choice of an Appropriate Information Dissimilarity Measure for Hierarchical Clustering of River Streamflow Time Series, Based on Calculated Lyapunov Exponent and Kolmogorov Measures. in Entropy Switzerland, Basel : MDPI., 21(2), 215. https://doi.org/10.3390/e21020215 https://hdl.handle.net/21.15107/rcub_dais_12914
Mihailović DT, Nikolić-Đorić E, Malinović-Milićević S, Singh VP, Mihailović A, Stošić T, Stošić B, Drešković N. The Choice of an Appropriate Information Dissimilarity Measure for Hierarchical Clustering of River Streamflow Time Series, Based on Calculated Lyapunov Exponent and Kolmogorov Measures. in Entropy. 2019;21(2):215. doi:10.3390/e21020215 https://hdl.handle.net/21.15107/rcub_dais_12914 .
Mihailović, Dragutin T., Nikolić-Đorić, Emilija, Malinović-Milićević, Slavica, Singh, Vijay P., Mihailović, Anja, Stošić, Tatjana, Stošić, Borko, Drešković, Nusret, "The Choice of an Appropriate Information Dissimilarity Measure for Hierarchical Clustering of River Streamflow Time Series, Based on Calculated Lyapunov Exponent and Kolmogorov Measures" in Entropy, 21, no. 2 (2019):215, https://doi.org/10.3390/e21020215 ., https://hdl.handle.net/21.15107/rcub_dais_12914 .