Primjer korištenja računalnog učenja u određivanju korelacije vodostaja, protoka i koncentracije nanosa

master's thesis

• Title (english): Example of using computer learning in determining the correlation of water level, flow and sediment concentration
• Author: Filip Đurkan
• Mentor: Hrvoje Meaški (mentor)
• Mentor: Dijana Oskoruš (pomoć pri izradi, neposredni voditelj)
• Committee member: Sanja Kovač (predsjednik povjerenstva)
• Committee member: Hrvoje Meaški (član povjerenstva)
• Committee member: Dijana Oskoruš (član povjerenstva)
• Committee member: Ranko Biondić (član povjerenstva)
• Committee member: Jelena Loborec (član povjerenstva)
• Granter: University of Zagreb; Faculty of Geotechnical Engineering; Varaždin
• Defense date and country: 2021-07-15, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Interdisciplinary Technical Sciences; Environmental Engineering

Abstract

Na temelju dobivenih podataka iz Državnog hidrometeorološkog zavoda za srednje dnevne protoke i vodostaje izmjerene na hidrološkim postajama Mursko Središće i Goričan na rijeci Muri napravljena je osnovna statistička analiza, empirijska i teoretska raspodjela podataka. Pomoću programskog jezika Python metodama računalnog učenja napravljena je korelacija protoka, vodostaja i koncentracije nanosa posebno za svaku postaju te za iste hidrološke veličine između dviju postaja. Korištene metode su: Linearna regresija, Polinomna regresija, Metoda potpornih vektora, Stablo odlučivanja, Algoritam slučajne šume. Svaka metoda se ocjenjuje pomoću određenih pokazatelja a u radu su korišteni: koeficijent determinacije R2, srednja kvadratna pogreška (MSE), srednja apsolutna pogreška (MAE), korijen srednje kvadratne pogreške (RMSE). Kada bi ocjena pokazatelja korelacije bila zadovoljavajuća, korelacije bi se moglo iskoristiti za nadopunjavanje podataka koji nedostaju ili koji nisu izmjereni.

Abstract (english)

Based on data given by Croatian meteorological and hydrological service for daily runoff and water levels measured on hydrological stations Mursko Središće and Goričan on river Mura statistical analysis, empirical and theoretical distributions were made. Machine Learning methods with the Python programming language were used to create correlation relationships of river runoff, water level and concentration of suspended solids. It was made for each station separately and between each station. Corelation methods used in the thesis are: Linear regression, Polynomial regression, Decision tree algorithm, Support vector regression and Random Forest algorithm. Every method is evaluated using following metrics: coefficient of determination R2, mean squared error (MSE), mean absolute error (MAE), root mean squared error (RMSE). If those parameters are acceptable, we could use correlations to fill missing data or the data that was not measured.

• Keywords: korelacija; statistika; hidrologija; računalno učenje; Python
• Keywords (english): correlation; statistics; hydrology; machine learning; Python
• Language: croatian
• URN:NBN: urn:nbn:hr:130:148075
• Study programme: Title: Geoenvironmental Engineering; specializations in: Geoenvironmental Engineering, Environmental Management, Water Management; Course: Water Management; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/ inženjerka inženjerstva okoliša (magistar/magistra inženjer/ inženjerka inženjerstva okoliša)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Access restricted to authenticated users
• Created on: 2021-10-29 07:08:07