Nebojša M. Ralević

Prof. Dr. Nebojša M. Ralević

orcid

Research Focus & Metrics

Academic Career Length

13Years (2014-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 February 2019 in Soft Computing
Verified
Journal Article
Closed access
PDF file
NRNebojša M. Ralević
Nebojša M. Ralević
DKDanijela Karaklić
Danijela Karaklić
NPNeda Pištinjat
Neda Pištinjat

10 February 2024 in Remote Sensing
Verified
Journal Article
Open access
PDF file
MPMarko Pavlovic
Marko Pavlovic
SISlobodan Ilic
Slobodan Ilic
NRNebojša M. Ralević
Nebojša M. Ralević
NANenad Antonic
Nenad Antonic
DRDylan Warren Raffa
Dylan Warren Raffa
DCDubravko Culibrk
Dubravko Culibrk
Monitoring soil organic carbon (SOC) typically assumes conducting a labor-intensive soil sampling campaign, followed by laboratory testing, which is both expensive and impractical for generating useful, spatially continuous data products. The present study leverages the power of machine learning (ML) and, in particular, deep neural networks (DNNs) for segmentation, as well as satellite imagery, to estimate the SOC remotely. We propose a new two-stage pipeline for remote SOC estimation, which relies on using a DNN trained to classify land cover to perform feature extraction, while the SOC estimation is performed by a different ML model. The first stage is an image segmentation DNN with the U-Net architecture, which is trained to estimate the land cover for an observed geographical region, based on multi-spectral images taken by the Sentinel-2 satellite constellation. This estimator is subsequently used to extract the latent feature vector for each of the output pixels, by rolling back from the output (dense) layer of the U-Net and accessing the last available convolutional layer of the same dimension as our desired output. The second stage is trained on a set of feature vectors extracted at the coordinates for which manual SOC measurements exist. We tested a variety of ML models and report on their performance. Using the best extremely randomized trees model, we generated a spatially continuous map of SOC estimations for the region of Tuscany, in Italy, with a resolution of 10 m, to share with the researchers as a means of validating the results and to demonstrate the efficiency of the proposed approach, which can can easily be scaled to create a global continuous SOC map.

9 July 2018 in Mathematical Problems in Engineering
Verified
Journal Article
Open access
PDF file
MPMarija Paunovic
Marija Paunovic
NRNebojša M. Ralević
Nebojša M. Ralević
VGVladimir Gajović
Vladimir Gajović
BVBiljana Mladenović Vojinović
Biljana Mladenović Vojinović
OMOlivera Milutinović
Olivera Milutinović
Dynamicity of the cloud computing system, complexity of real-world decision-making problems, and nontransparent market imply numerous difficulties in finding the corresponding solution for cloud supplier evaluation and selection. Due to that reason, there is a need to develop a model that can contribute to increase of reliability of evaluation and reduction of subjectivity of the decision-maker. Having in mind the aforementioned, the authors proposed a new concept to model a cloud supplier evaluation based on technology perspective. Here, we developed a model that combined fuzzy techniques with implemented simulation of model parameters. The proposed method for cloud supplier evaluation is flexible with an objective evaluation and ensures more precise numerical value. Numerical examples are presented to demonstrate the effectiveness and pragmatism of the proposed approach.

1 January 2020 in Mathematica Montisnigri
Verified
Journal Article
Closed access
NRNebojša M. Ralević
Nebojša M. Ralević
MPMarija Paunovic
Marija Paunovic
Bratislav IricaninBratislav Iricanin
Bratislav Iricanin

1 January 2014 in Journal of the Serbian Chemical Society
Verified
Journal Article
Open access
AMAleksandra Mihailovic
Aleksandra Mihailovic
VVVucinic Vasic
Vucinic Vasic
JNJordana Ninkov
Jordana Ninkov
SESuzana Eric
Suzana Eric
NRNebojša M. Ralević
Nebojša M. Ralević
AAAleksandar Antić
Aleksandar Antić
During December 2009 snow was collected at twenty two locations across the urban area of Novi Sad, directly from roads and from traffic islands near crossroads. The total metal concentration was determined for each of ten metals (Al, Ca, Cu, Fe, K, Mn, Na, Ni, Pb, Zn) using the ICP-OES analytical technique. The metal found to have the lowest concentration was Ni (0.0265 mg dm-3). The metal with the highest concentration was Na (10786 mg dm-3), which is a consequence of sodium chloride being used as de-icing salt on the roads. The metal with the second highest concentration at all locations was Ca; this was most likely the result of soil dust. The Spearman's rank correlation coefficients between analyzed metals were calculated to determine how concentrations of metals were related. A cluster analysis was carried out on the obtained data sets, using both hierarchical and partitioning methods in order to identify associations among metals and/or locations. It was shown that traffic density was not the most important factor that caused differences between the concentrations of the metals in samples.

1 February 2021 in Information Sciences
Verified
Journal Article
Closed access
VIVladimir Ilić
Vladimir Ilić
NRNebojša M. Ralević
Nebojša M. Ralević