Mario Dudjak

@ferit.unios.hr

Postdoctoral Researcher at Department of Software Engineering
Faculty of Electrical Engineering, Computer Science and Information Technology Osijek

Mario Dudjak


                        Download Compact PDF  
https://researchid.co/mariodudjak

EDUCATION

PhD in Computer Science (Artificial Intelligence)
Master of Computer Science

RESEARCH, TEACHING, or OTHER INTERESTS

Computer Science, Artificial Intelligence
15

Scopus Publications

270

Scholar Citations

7

Scholar h-index

6

Scholar i10-index

Scopus Publications

  • Reliable Authentication of DRAM PUFs by Leveraging Random Forests in Machine Learning
    Mario Dudjak, Mariam Elsharif, Shekoufeh Neisarian, Tolga Arul, Stefan Katzenbeisser, Nikolaos Athanasios Anagnostopoulos, Nico Mexis
    Proceedings of the IEEE World Forum on Internet of Things Wf Iot, 2025
    The growing demand for smaller, lighter, and more embedded hardware has made Physical Unclonable Functions (PUFs) a promising solution for authentication in Internet of Things (IoT) applications. Traditional PUF authentication methods often rely on Error Correction (EC), which can be computationally intensive and time-consuming. Given the time-critical nature of authentication in security-sensitive IoT systems, there is a need for efficient alternatives. Machine Learning (ML) techniques have emerged as a viable option for bypassing EC by leveraging pre-trained models. This paper proposes a novel approach for authenticating DRAM PUFs using Random Forest (RF) classifiers. RFs, which are ensembles of independently trained decision trees, leverage collective voting to improve classification accuracy. This approach enhances confidence, reduces bias, and offers greater transparency, as the decision-making process of decision trees is inherently interpretable. Our results demonstrate that RF classifiers provide robust performance, comparable to or superior to other ML approaches commonly employed for these tasks, offering a reliable and efficient alternative to EC-based methods.
  • Reducing Meta-Level Data in Stacking Ensembles for Classification
    Dražen Bajer, Bruno Zorić, Mario Dudjak
    Proceedings of International Conference on Smart Systems and Technologies Sst 2024, 2024
    Stacking as an ensemble framework relies on meta-learning for combining the outputs of multiple base classifiers. What exactly the outputs represent and how they are used for meta-learning is an important aspect of stacking. Commonly, the prediction probabilities produced by the base classifiers at the class level are used for training the meta-classifier. The dimensionality of this meta-level data is, therefore, not negligible. Reducing it would result in the need for a less complex meta-classifier, and would thereby simplify the process of meta-learning. This paper presents a straightforward approach for aggregating the prediction probabilities of the base classifiers. By calculating the mean and standard deviation at the class level, a substantial reduction in the dimensionality of the meta-level data is achieved. The experimental analysis, conducted on multiple diverse datasets, suggests that the aggregation preserves sufficient information for meta-learning since better or highly competitive performance was attained with respect to the common stacking ensemble framework (utilising the full meta-level data).
  • Feature extraction procedures for chronic wound tissue classification
    Bruno Zorić, Dražen Bajer, Mario Dudjak
    2024 IEEE Zooming Innovation in Consumer Technologies Conference Zinc 2024, 2024
    As healthcare services become both more necessary and expensive on a daily basis, the possibility arises for the inclusion of (semi)automated machine learning based approaches for aiding healthcare professionals and, hopefully, reducing costs. The area of chronic wound management, which includes diagnostics, healing process tracking, treatment related decision-making etc., represents a prominent candidate for such applications. This paper considers a straightforward approach to the classification of different tissue types constituting chronic wounds. It delves into the merits of several different feature extraction procedures and their descriptive properties when utilised at the pixel level, with the goal of selecting simple descriptors yielding solid performance and robust model behaviour. A multifaceted comparative analysis was conducted which considered performance on data stemming from both a prosthetic model and real patients in terms of different feature extraction procedures, colour spaces and window sizes. Even though the best advice anyone can ever offer is to adapt to the data at hand, the results do point at some possible shortcuts regarding feature extraction that lead to overall good performance figures.
  • An ensemble-based framework for biomedical classification problems
    Mario Dudjak, Bruno Zorić, Dražen Bajer
    2023 IEEE Zooming Innovation in Consumer Technologies Conference Zinc 2023, 2023
    Model selection is an essential step when applying machine learning to classification problems. It is typically carried out by the practitioner who strives to identify the most suitable classifier for a given problem. Given the variety of classifiers available and the difficulty in predicting which one will yield the best performance depending on the characteristics of the problem, this is by no means a simple task. Biomedical problems pose a significant challenge in this regard due to their numerous data intrinsic characteristics that are known to impair classification performance. Given that different classifiers perform well for different biomedical problems, combining them into an ensemble would seem practical. However, the practitioner still needs to determine how to combine them. This paper presents an ensemble-based framework that automates the training and combination of different classifiers in order to relieve practitioners of this burden whilst obtaining highly competitive performance. The effectiveness of the proposed framework was evaluated on several biomedical problems from the literature.
  • Predicting public transport arrival time and congestion based on BLE beacon crowdsourced data
    Bruno Zoric, Mario Dudjak, Drazen Bajer
    2022 IEEE Zooming Innovation in Consumer Technologies Conference Zinc 2022, 2022
    Public transport networks play an important role in minimising congestion and improving environmental sustainability of developed cities. However, they face a number of challenges in achieving these goals, especially during the ongoing pandemic. In order to overcome these challenges, at least to some extent, public transport must be made accessible and attractive to potential passengers. To this end, a design for augmenting the transit network is proposed in this paper. The utilisation of Bluetooth low energy beacons as one of the key components makes it cost-effective and easily applicable in such an environment. Additionally, it incorporates a simple mobile application used to enable crowdsourced data acquisition on which machine learning-based models can be built to predict information relevant to consumers, like arrival time and congestion estimates. A prototype of the proposed system design, albeit of limited functionality, was deployed and evaluated on a tram route in the city of Osijek, Croatia. Promising results were obtained in terms of congestion and arrival time prediction, but some challenges remain to be addressed, like motivating users to participate in the crowdsourced data collection.
  • SMOTE Inspired Extension for Differential Evolution
    Dražen Bajer, Bruno Zorić, Mario Dudjak
    Lecture Notes in Computer Science Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics, 2022
  • Bio-inspired wrapper-based feature selection: does the choice of metric matter?
    Dracen Bajer, Mario Dudjak, Bruno Zoric
    Proceedings of International Conference on Smart Systems and Technologies Sst 2022, 2022
    Applications of bio-inspired algorithms to feature selection offering promising performance can be frequently encountered in the literature. Serving as wrappers, their application essentially boils down to the selection of a classifier and an appropriate metric for evaluating model performance. Although the choice of classifier was investigated to a viable extent, this is, however, not the case for the choice of metric and its impact on performance. Nevertheless, the utilisation of a wide variety of metrics is clearly apparent, albeit to different proportions. This raises the question of whether some metrics might be a better choice than others for bio-inspired wrapper-based methods. This paper sheds some light on this matter by comparing different metrics from different perspectives, like correlation, general performance and subset sizes. Unexpected results were obtained, to say the least, apart from some minor exceptions. Generally, no single metric appears to have an edge over the others. Yet, further investigation is certainly warranted.
  • An empirical study of data intrinsic characteristics that make learning from imbalanced data difficult[Formula presented]
    Mario Dudjak, Goran Martinović
    Expert Systems with Applications, 2021
  • Wrapper-based feature selection via differential evolution: Benchmarking different discretisation techniques
    Bruno Zoric, Drazen Bajer, Mario Dudjak
    Proceedings of 2020 International Conference on Smart Systems and Technologies Sst 2020, 2020
    Wrapper-based feature selection approaches reliant on different bio-inspired optimisation algorithms are both effective and widely employed when dealing with classification problems. These algorithms have proven themselves as successful wrappers in finding good feature subsets. However, as a large number of them is defined for the real domain, the small detail of their adaptation to the discrete domain of feature selection is often overlooked. This holds especially true for differential evolution, a prominent wrapper choice among bioinspired optimisation algorithms. As distinct discretisation techniques have been proposed in the literature, the question of which one to incorporate in differential evolution and under which circumstances remains rather unanswered. This paper attempts to provide some answers in that regard by studying the incorporation of discretisation techniques into differential evolution and their influence on the quality of attained feature subsets. Given their differences, some suggestions concerning the selection of discretisation techniques are given based on the obtained results.
  • Wrapper-based feature selection: How important is the wrapped classifier?
    Drazen Bajer, Mario Dudjak, Bruno Zoric
    Proceedings of 2020 International Conference on Smart Systems and Technologies Sst 2020, 2020
    Wrapper-based feature (subset) selection is a frequently used approach for dataset dimensionality reduction, especially when dealing with classification problems. The choice of wrapper is at the forefront of these approaches, whilst the choice of the classifier is typically based on its simplicity as to reduce the computational cost. Since the search is guided by the selected classifier, the same one is also later used for independent testing. This raises the question of how well such feature subsets are suited for other types of classifiers. In other words, can one classifier be used for finding feature subsets that are also effective for others? An investigation into this matter was performed by testing and analysing the utility of subsets found by one classifier with respect to other classifiers. It hints at the importance of classifier choice since some models, whilst used inside the wrapper, can solely conform the dataset to themselves, whilst others are less susceptible to this issue. Consequently, an insight into the robustness of the employed classifiers was gained as well.
  • In-depth performance analysis of SMOTE-based oversampling algorithms in binary classification
    International Journal of Electrical and Computer Engineering Systems, 2020
  • An API-first methodology for designing a microservice-based backend as a service platform
    Mario Dudjak, Goran Martinović
    Information Technology and Control, 2020
  • Evaluation and analysis of bio-inspired optimization algorithms for feature selection
    Drazen Bajer, Bruno Zoric, Mario Dudjak, Goran Martinovic
    Informatics 2019 IEEE 15th International Scientific Conference on Informatics Proceedings, 2019
  • Performance Analysis of SMOTE-based Oversampling Techniques When Dealing with Data Imbalance
    Drazen Bajer, Bruno Zonc, Mario Dudjak, Goran Martinovic
    International Conference on Systems Signals and Image Processing, 2019
  • Design and development of a smart attendance management system with Bluetooth low energy beacons
    Bruno Zoric, Mario Dudjak, Drazen Bajer, Goran Martinovic
    2019 Zooming Innovation in Consumer Technologies Conference Zinc 2019, 2019

RECENT SCHOLAR PUBLICATIONS

  • Reliable Authentication of DRAM PUFs by Leveraging Random Forests in Machine Learning
    M Dudjak, M Elsharif, S Neisarian, T Arul, S Katzenbeisser, ...
    2025 IEEE 11th World Forum on Internet of Things (WF-IoT), 1-7 , 2025
    2025
    Citations: 1
  • Empirijska studija postupaka strojnog učenja za prepoznavanje malicioznih napada
    A Carević, M Dudjak
    Festung: Časopis za interdisciplinarna istraživanja u poslovanju 1 (1), 15-24 , 2025
    2025
  • An Empirical Study of Machine Learning Techniques for Malicious Attack Detection
    A Carević, M Dudjak
    Festung: Časopis za interdisciplinarna istraživanja u poslovanju 1 (1), 15-24 , 2025
    2025
  • Reducing Meta-Level Data in Stacking Ensembles for Classification
    D Bajer, B Zorić, M Dudjak
    2024 International Conference on Smart Systems and Technologies (SST), 133-138 , 2024
    2024
    Citations: 1
  • Feature extraction procedures for chronic wound tissue classification
    B Zorić, D Bajer, M Dudjak
    2024 Zooming Innovation in Consumer Technologies Conference (ZINC), 1-6 , 2024
    2024
    Citations: 1
  • An empirical study of the crossover operator in a genetic algorithm used as a wrapper for feature selection
    M Dudjak
    Authorea Preprints , 2023
    2023
  • An ensemble-based framework for biomedical classification problems
    M Dudjak, B Zorić, D Bajer
    2023 Zooming Innovation in Consumer Technologies Conference (ZINC), 69-73 , 2023
    2023
    Citations: 3
  • SMOTE Inspired Extension for Differential Evolution
    D Bajer, B Zorić, M Dudjak
    International Conference on Bioinspired Optimization Methods and Their … , 2022
    2022
  • Bio-inspired wrapper-based feature selection: does the choice of metric matter?
    D Bajer, M Dudjak, B Zorić
    2022 International Conference on Smart Systems and Technologies (SST), 1-8 , 2022
    2022
  • Učenje iz neuravnoteženih podataka unaprijeđenim postupcima za odabir značajki, preuzorkovanje i izgradnju radijalnih neuronskih mreža
    M Dudjak
    Sveučilište Josipa Jurja Strossmayera u Osijeku, Sveučilište Josipa Jurja … , 2022
    2022
    Citations: 5
  • Predicting public transport arrival time and congestion based on BLE beacon crowdsourced data
    B Zorić, M Dudjak, D Bajer
    2022 IEEE Zooming Innovation in Consumer Technologies Conference (ZINC), 81-86 , 2022
    2022
    Citations: 3
  • An empirical study of data intrinsic characteristics that make learning from imbalanced data difficult
    M Dudjak, G Martinović
    Expert systems with applications 182, 115297 , 2021
    2021
    Citations: 54
  • Wrapper-based feature selection: how important is the wrapped classifier?
    D Bajer, M Dudjak, B Zorić
    2020 International conference on smart systems and technologies (SST), 97-105 , 2020
    2020
    Citations: 30
  • Wrapper-based feature selection via differential evolution: benchmarking different discretisation techniques
    B Zorić, D Bajer, M Dudjak
    2020 International Conference on Smart Systems and Technologies (SST), 89-96 , 2020
    2020
    Citations: 6
  • An API-first methodology for designing a microservice-based Backend as a Service platform
    M Dudjak, G Martinović
    Information technology and control 49 (2), 206-223 , 2020
    2020
    Citations: 60
  • In-Depth Performance Analysis of SMOTE-Based Oversampling Algorithms in Binary Classification
    M Dudjak, G Martinović
    International Journal of Electrical and Computer Engineering Systems 11 (1 … , 2020
    2020
    Citations: 20
  • An empirical study of classification algorithms when dealing with the problem of class imbalance and other data intrinsic characteristics
    M Dudjak, G Martinović
    Abstract Book-Fifth International Workshop on Data Science/Lončarić, Sven … , 2020
    2020
  • Benchmarking bio-inspired computation algorithms as wrappers for feature selection
    D Bajer, B Zorić, M Dudjak, G Martinović
    Acta electrotechnica et informatica 20 (2), 35-43 , 2020
    2020
    Citations: 4
  • Evaluation and analysis of bio-inspired optimization algorithms for feature selection
    D Bajer, B Zorić, M Dudjak, G Martinović
    2019 IEEE 15th international scientific conference on informatics, 000285-000292 , 2019
    2019
    Citations: 7
  • Performance analysis of SMOTE-based oversampling techniques when dealing with data imbalance
    D Bajer, B Zonć, M Dudjak, G Martinović
    2019 International Conference on Systems, Signals and Image Processing … , 2019
    2019
    Citations: 47

MOST CITED SCHOLAR PUBLICATIONS

  • An API-first methodology for designing a microservice-based Backend as a Service platform
    M Dudjak, G Martinović
    Information technology and control 49 (2), 206-223 , 2020
    2020
    Citations: 60
  • An empirical study of data intrinsic characteristics that make learning from imbalanced data difficult
    M Dudjak, G Martinović
    Expert systems with applications 182, 115297 , 2021
    2021
    Citations: 54
  • Performance analysis of SMOTE-based oversampling techniques when dealing with data imbalance
    D Bajer, B Zonć, M Dudjak, G Martinović
    2019 International Conference on Systems, Signals and Image Processing … , 2019
    2019
    Citations: 47
  • Wrapper-based feature selection: how important is the wrapped classifier?
    D Bajer, M Dudjak, B Zorić
    2020 International conference on smart systems and technologies (SST), 97-105 , 2020
    2020
    Citations: 30
  • Design and development of a smart attendance management system with Bluetooth low energy beacons
    B Zorić, M Dudjak, D Bajer, G Martinović
    2019 Zooming Innovation in Consumer Technologies Conference (ZINC), 86-91 , 2019
    2019
    Citations: 21
  • In-Depth Performance Analysis of SMOTE-Based Oversampling Algorithms in Binary Classification
    M Dudjak, G Martinović
    International Journal of Electrical and Computer Engineering Systems 11 (1 … , 2020
    2020
    Citations: 20
  • Evaluation and analysis of bio-inspired optimization algorithms for feature selection
    D Bajer, B Zorić, M Dudjak, G Martinović
    2019 IEEE 15th international scientific conference on informatics, 000285-000292 , 2019
    2019
    Citations: 7
  • Wrapper-based feature selection via differential evolution: benchmarking different discretisation techniques
    B Zorić, D Bajer, M Dudjak
    2020 International Conference on Smart Systems and Technologies (SST), 89-96 , 2020
    2020
    Citations: 6
  • Survey of database backup management
    M Dudjak, I Lukić, M Köhler
    27th International Scientific and Professional Conference Organization and … , 2017
    2017
    Citations: 6
  • Učenje iz neuravnoteženih podataka unaprijeđenim postupcima za odabir značajki, preuzorkovanje i izgradnju radijalnih neuronskih mreža
    M Dudjak
    Sveučilište Josipa Jurja Strossmayera u Osijeku, Sveučilište Josipa Jurja … , 2022
    2022
    Citations: 5
  • Benchmarking bio-inspired computation algorithms as wrappers for feature selection
    D Bajer, B Zorić, M Dudjak, G Martinović
    Acta electrotechnica et informatica 20 (2), 35-43 , 2020
    2020
    Citations: 4
  • An ensemble-based framework for biomedical classification problems
    M Dudjak, B Zorić, D Bajer
    2023 Zooming Innovation in Consumer Technologies Conference (ZINC), 69-73 , 2023
    2023
    Citations: 3
  • Predicting public transport arrival time and congestion based on BLE beacon crowdsourced data
    B Zorić, M Dudjak, D Bajer
    2022 IEEE Zooming Innovation in Consumer Technologies Conference (ZINC), 81-86 , 2022
    2022
    Citations: 3
  • Reliable Authentication of DRAM PUFs by Leveraging Random Forests in Machine Learning
    M Dudjak, M Elsharif, S Neisarian, T Arul, S Katzenbeisser, ...
    2025 IEEE 11th World Forum on Internet of Things (WF-IoT), 1-7 , 2025
    2025
    Citations: 1
  • Reducing Meta-Level Data in Stacking Ensembles for Classification
    D Bajer, B Zorić, M Dudjak
    2024 International Conference on Smart Systems and Technologies (SST), 133-138 , 2024
    2024
    Citations: 1
  • Feature extraction procedures for chronic wound tissue classification
    B Zorić, D Bajer, M Dudjak
    2024 Zooming Innovation in Consumer Technologies Conference (ZINC), 1-6 , 2024
    2024
    Citations: 1
  • Izrada BaaS (Backend as a Service) sustava za web aplikacije
    M Dudjak
    Sveučilište Josipa Jurja Strossmayera u Osijeku, Sveučilište Josipa Jurja … , 2018
    2018
    Citations: 1
  • Empirijska studija postupaka strojnog učenja za prepoznavanje malicioznih napada
    A Carević, M Dudjak
    Festung: Časopis za interdisciplinarna istraživanja u poslovanju 1 (1), 15-24 , 2025
    2025
  • An Empirical Study of Machine Learning Techniques for Malicious Attack Detection
    A Carević, M Dudjak
    Festung: Časopis za interdisciplinarna istraživanja u poslovanju 1 (1), 15-24 , 2025
    2025
  • An empirical study of the crossover operator in a genetic algorithm used as a wrapper for feature selection
    M Dudjak
    Authorea Preprints , 2023
    2023