Multilevel split of high-dimensional water quality data using artificial neural networks for the prediction of dissolved oxygen in the Danube River
Samo za registrovane korisnike
2020
Članak u časopisu (Objavljena verzija)
Metapodaci
Prikaz svih podataka o dokumentuApstrakt
In this study, a self-organizing network-based monitoring location similarity index (LSI) was coupled with Ward neural networks (WNNs) with the aim to create a more accurate, but less complex, multiple sites model for the prediction of dissolved oxygen (DO) content. This multilevel splitting approach comprises the LSI-based grouping of monitoring locations according to their similarity, and virtual splitting of processed data based on their features using WNN. The values of 18 water quality parameters monitored for 12 years at 17 sites on the Danube River flow thought Serbia were used. The optimal input combinations were selected using partial mutual information algorithm with termination based on the Akaike information criterion. LSI-based splitting has yielded two groups of monitoring sites that were modeled with separate WNN models. The number and types of selected inputs differed between those two groups of sites, which was in agreement with possible pollution sources. Multiple per...formance metrics have revealed that the WNN models perform similar or better than multisite DO prediction models published in the literature, while using two to four times less inputs and data patterns.
Ključne reči:
Similarity metrics / PMIS / DO prediction / Ward neural networkIzvor:
Neural Computing & Applications, 2020, 32, 8, 3957-3966Izdavač:
- Springer London Ltd, London
Finansiranje / projekti:
DOI: 10.1007/s00521-019-04079-y
ISSN: 0941-0643
WoS: 000524416400061
Scopus: 2-s2.0-85062026994
Institucija/grupa
Tehnološko-metalurški fakultetTY - JOUR AU - Antanasijević, Davor AU - Pocajt, Viktor AU - Perić-Grujić, Aleksandra AU - Ristić, Mirjana PY - 2020 UR - http://TechnoRep.tmf.bg.ac.rs/handle/123456789/4574 AB - In this study, a self-organizing network-based monitoring location similarity index (LSI) was coupled with Ward neural networks (WNNs) with the aim to create a more accurate, but less complex, multiple sites model for the prediction of dissolved oxygen (DO) content. This multilevel splitting approach comprises the LSI-based grouping of monitoring locations according to their similarity, and virtual splitting of processed data based on their features using WNN. The values of 18 water quality parameters monitored for 12 years at 17 sites on the Danube River flow thought Serbia were used. The optimal input combinations were selected using partial mutual information algorithm with termination based on the Akaike information criterion. LSI-based splitting has yielded two groups of monitoring sites that were modeled with separate WNN models. The number and types of selected inputs differed between those two groups of sites, which was in agreement with possible pollution sources. Multiple performance metrics have revealed that the WNN models perform similar or better than multisite DO prediction models published in the literature, while using two to four times less inputs and data patterns. PB - Springer London Ltd, London T2 - Neural Computing & Applications T1 - Multilevel split of high-dimensional water quality data using artificial neural networks for the prediction of dissolved oxygen in the Danube River EP - 3966 IS - 8 SP - 3957 VL - 32 DO - 10.1007/s00521-019-04079-y ER -
@article{ author = "Antanasijević, Davor and Pocajt, Viktor and Perić-Grujić, Aleksandra and Ristić, Mirjana", year = "2020", abstract = "In this study, a self-organizing network-based monitoring location similarity index (LSI) was coupled with Ward neural networks (WNNs) with the aim to create a more accurate, but less complex, multiple sites model for the prediction of dissolved oxygen (DO) content. This multilevel splitting approach comprises the LSI-based grouping of monitoring locations according to their similarity, and virtual splitting of processed data based on their features using WNN. The values of 18 water quality parameters monitored for 12 years at 17 sites on the Danube River flow thought Serbia were used. The optimal input combinations were selected using partial mutual information algorithm with termination based on the Akaike information criterion. LSI-based splitting has yielded two groups of monitoring sites that were modeled with separate WNN models. The number and types of selected inputs differed between those two groups of sites, which was in agreement with possible pollution sources. Multiple performance metrics have revealed that the WNN models perform similar or better than multisite DO prediction models published in the literature, while using two to four times less inputs and data patterns.", publisher = "Springer London Ltd, London", journal = "Neural Computing & Applications", title = "Multilevel split of high-dimensional water quality data using artificial neural networks for the prediction of dissolved oxygen in the Danube River", pages = "3966-3957", number = "8", volume = "32", doi = "10.1007/s00521-019-04079-y" }
Antanasijević, D., Pocajt, V., Perić-Grujić, A.,& Ristić, M.. (2020). Multilevel split of high-dimensional water quality data using artificial neural networks for the prediction of dissolved oxygen in the Danube River. in Neural Computing & Applications Springer London Ltd, London., 32(8), 3957-3966. https://doi.org/10.1007/s00521-019-04079-y
Antanasijević D, Pocajt V, Perić-Grujić A, Ristić M. Multilevel split of high-dimensional water quality data using artificial neural networks for the prediction of dissolved oxygen in the Danube River. in Neural Computing & Applications. 2020;32(8):3957-3966. doi:10.1007/s00521-019-04079-y .
Antanasijević, Davor, Pocajt, Viktor, Perić-Grujić, Aleksandra, Ristić, Mirjana, "Multilevel split of high-dimensional water quality data using artificial neural networks for the prediction of dissolved oxygen in the Danube River" in Neural Computing & Applications, 32, no. 8 (2020):3957-3966, https://doi.org/10.1007/s00521-019-04079-y . .