Predrag S. Petronijević

Dr. Predrag S. Petronijević

orcid

Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

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

Representative Publications

View All Publications
17 January 2025 in Computation
Verified
Journal Article
Open access
PDF file
Miljan KovačevićMiljan Kovačević
Miljan Kovačević
Marijana Hadzima-NyarkoMarijana Hadzima-Nyarko
Marijana Hadzima-Nyarko
PPPredrag S. Petronijević
Predrag S. Petronijević
TVTatijana Vasiljević
Tatijana Vasiljević
MRMiroslav Radomirović
Miroslav Radomirović
This study presents a detailed analysis of various machine learning models for predicting the interfacial bond strength of fiber-reinforced polymer (FRP) concrete, including multiple linear regression, Multigene Genetic Programming (MGGP), an ensemble of regression trees, Gaussian Process Regression (GPR), Support Vector Regression (SVR), and neural networks. The evaluation was based on their predictive accuracy. The optimal model identified was the GPR ARD Exponential model, which achieved a mean absolute error (MAE) of 1.8953 MPa and a correlation coefficient (R) of 0.9658. An analysis of this optimal model highlighted the most influential variables affecting the bond strength. Additionally, the research identified several models with lower expression complexity and reduced accuracy, which may still be applicable in practical scenarios.

2 March 2026 in Applied Sciences
Verified
Journal Article
Open access
PDF file
Miljan KovačevićMiljan Kovačević
Miljan Kovačević
ACAnđelko Crnoja
Anđelko Crnoja
Borko BulajicBorko Bulajic
Borko Bulajic
PPPredrag S. Petronijević
Predrag S. Petronijević
The increasing accumulation of end-of-life tires has motivated the development of sustainable construction materials incorporating recycled rubber for acoustic insulation applications. This study proposes a data-driven framework for predicting the weighted airborne sound reduction index (Rw) of recycled rubber–polyurethane composite panels based on a limited experimental dataset. Specimens with varying granulometric composition, material density, and polyurethane adhesive dosage were evaluated in accordance with EN ISO 10140-2:2010 and EN ISO 717-1:2013. To address data scarcity, a regression-oriented SMOTE strategy was applied exclusively to the training set to preserve statistical representativeness and avoid data leakage. Test set representativeness was ensured by systematically evaluating numerous data splits and adopting the one that maximized multivariate statistical consistency. A hierarchical modeling approach was adopted, ranging from classical regression models to tree-based ensemble methods and multigene symbolic regression. Model performance was evaluated using R2, RMSE, MAE, and MAPE on an independent test set. The highest accuracy and robustness were obtained using symbolic regression, with R2 values close to 0.99 and minimal prediction errors. Shapley value analysis and PDP/ICE plots identified material density as the dominant predictor of Rw, followed by polyurethane adhesive dosage, while granulometric composition exhibited a weaker influence. The proposed framework provides an accurate and interpretable tool for the preliminary design and optimization of recycled rubber acoustic panels.

1 December 2021
Verified
Conference Paper
Open access
PDF file
NSNevena Simić
Nevena Simić
PPPredrag S. Petronijević
Predrag S. Petronijević
ADAleksandar Devedžić
Aleksandar Devedžić
MIMarija Zoran Ivanovic
Marija Zoran Ivanovic
This paper analyses the problem of estimating the required quantities of major work items in the construction of residential and residential-commercial buildings using machine learning algorithms. The goal is to form a model that will provide a fast and sufficiently accurate estimate of the quantities of major work items, based on a small amount of known information on the technical characteristics and the environment of future residential and residential-commercial buildings. The case study included 71 projects of residential and residential-commercial buildings construction realised on the territory of the Republic of Serbia. Several models have been developed, and the paper presents those models that had the best performances. The models developed in this way can significantly contribute to resource planning and the accuracy of cost estimates in the early project phases.

1 January 2026
Verified
Conference Paper
Closed access
IMIvan Milojković
Ivan Milojković
NINenad Ivanišević
Nenad Ivanišević
PPPredrag S. Petronijević
Predrag S. Petronijević
Ekspertne metode se koriste za predlaganje i nalaženje rešenja složenih i slabo strukturiranih problema, čiji se parametri ne mogu odrediti na osnovu statističkih ili eksperimentalnih podataka. Ove metode baziraju se na subjektivnim ocenama eksperata za problem koji treba rešiti i među njima se posebno ističe Delfi metoda. U radu se daje pregled primene Delfi metode u odnosu na hidrotehnički objekat – kanalizacionu crpnu stanicu. Dobijaju se ocene kriterijuma po kojima se alternative vrednuju.

25 November 2024 in Preprints.org
Verified
Preprint
Open access
PDF file
Miljan KovačevićMiljan Kovačević
Miljan Kovačević
Marijana Hadzima-NyarkoMarijana Hadzima-Nyarko
Marijana Hadzima-Nyarko
PPPredrag S. Petronijević
Predrag S. Petronijević
TVTatijana Vasiljević
Tatijana Vasiljević
MRMiroslav Radomirović
Miroslav Radomirović
This study presents a detailed analysis of various machine learning models for predicting the interfacial bond strength of fiber-reinforced polymer (FRP) concrete, utilizing multiple linear regression, an ensemble of regression trees, Gaussian Process Regression (GPR), Support Vector Regression (SVR), Multigene Genetic Programming (MGGP), and neural network models. The models were evaluated based on their accuracy. The optimal model identified was the GPR ARD model, which achieved a mean absolute error (MAE) of 1.8953 MPa and a correlation coefficient (R) of 0.9658. Analysis of this optimal model highlighted that the three most influential variables affecting the bond strength are the length of the FRP strip (Lf), the thickness of the strip ( tf), and the compressive strength of the concrete to which the strip is applied (fc'). Additionally, the research identified several models with lower expression complexity and reduced accuracy, which may still be applicable in practical scenarios.

29 October 2026
Verified
Conference Paper
Closed access
PPPredrag S. Petronijević
Predrag S. Petronijević
NINenad Ivanišević
Nenad Ivanišević
ŽMŽeljko Marinković
Željko Marinković
NKNikola Knežević
Nikola Knežević
In contemporary construction management, the Critical Path Method serves as a primary contractual instrument for tracking progress and resolving delay claims. However, the reliability of these models depends heavily on the structural integrity of their network logic. This review paper provides a critical synthesis of global standards used to evaluate dynamic construction schedule quality. Focusing on established frameworks like the DCMA 14-Point Assessment and GAO guidelines, the study analyses the engineering rationales behind quantitative algorithmic thresholds, including open ends, invalid constraints, and lag misuse. Additionally, we examine software anomalies encountered during progress updates and discuss conformance scoring in contract administration. By bridging theoretical network math and empirical site constraints, this review outlines industry best practices and identifies future research directions for testing automated compliance tools.