Miodrag Djordjevic
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Miodrag Djordjevic has a total of 44 co-authors. Here are the top co-authors by number of shared publications:
- Aleksandar S. Nastic
Faculty of Sciences and Mathematics, University of Niš, Niš, Serbia
3 shared publications - Ana Savić
Department of Biology and Ecology, Faculty of Sciences and Mathematics, University of Niš, Višegradska 33, 18000 Niš, Serbia
2 shared publications - Danijela Nikolić
Department of Biology and Ecology, Faculty of Sciences and Mathematics, University of Niš, Niš, Serbia
2 shared publications - Dragana Jenackovic-Gocic
Department of Biology and Ecology, Faculty of Sciences and Mathematics, University of Niš, Niš, Serbia
2 shared publications - Irena Raca
Department of Biology and Ecology, Faculty of Sciences and Mathematics, University of Niš, Niš, Serbia
2 shared publications - Marina Jušković
Department of Biology and Ecology, Faculty of Sciences and Mathematics, University of Niš, Niš, Serbia
2 shared publications
Latest Publications
- Open Access
- 30 September 2026
- Published by University of Nis in Facta Universitatis, Series: Mathematics and Informatics
- p. 489-506
THE CORRELATION STRUCTURE OF THE LATENT COMPONENTS OF THE TIME SERIES WITH DISCRETE LAPLACE MARGINALS AND ITS IMPACT ON PARAMETER ESTIMATION
The main subject of this paper is the proposal of a new method for parameter estimation, applicable to time series with a discrete Laplace marginal distribution. The proposed estimators are derived from the correlation structure of the latent components of the time series. In this context, the complete correlation structure of the latent components is established and thoroughly analyzed. Based on these results, new estimators are defined, and their key statistical properties, including unbiasedness, consistency, and asymptotic behavior, are rigorously examined. Owing to their fast convergence properties, the proposed estimators are particularly advantageous in low-data regimes, enabling, for example, more effective training of machine learning models on limited datasets. A comparison with existing estimators is also provided, highlighting several important improvements in efficiency and applicability. Finally, all theoretical findings are supported and validated through simulation studies and real-life data applications.
- 30 September 2026
- Published by University of Nis in Facta Universitatis, Series: Mathematics and Informatics
- p. 489-506
Open AccessTHE CORRELATION STRUCTURE OF THE LATENT COMPONENTS OF THE TIME SERIES WITH DISCRETE LAPLACE MARGINALS AND ITS IMPACT ON PARAMETER ESTIMATION
The main subject of this paper is the proposal of a new method for parameter estimation, applicable to time series with a discrete Laplace marginal distribution. The proposed estimators are derived from the correlation structure of the latent components of the time series. In this context, the complete correlation structure of the latent components is established and thoroughly analyzed. Based on these results, new estimators are defined, and their key statistical properties, including unbiasedness, consistency, and asymptotic behavior, are rigorously examined. Owing to their fast convergence properties, the proposed estimators are particularly advantageous in low-data regimes, enabling, for example, more effective training of machine learning models on limited datasets. A comparison with existing estimators is also provided, highlighting several important improvements in efficiency and applicability. Finally, all theoretical findings are supported and validated through simulation studies and real-life data applications.
- Open Access
- 1 January 2025
- Published by National Library of Serbia in Filomat
- Vol. 39 (21) , 7173-7193
Value-correlation controlled integer-valued autoregressive process in random environment
The manuscript presents a new integer-valued first-order autoregressive process in a random environment, which is governed by two control processes. The first process defines the marginal distribu-tion, while the second regulates the correlation structure within the model. The properties of the proposed model are examined in detail, providing insights into its theoretical foundations and practical implications. Two methods for estimating the unknown parameters are introduced: the Yule-Walker estimator and the conditional maximum likelihood estimator. A series of simulations assesses the efficiency of these esti-mation techniques and demonstrates their performance across various scenarios. The effectiveness of the introduced model is further evaluated through its application to real-life data.
- 1 January 2025
- Published by National Library of Serbia in Filomat
- Vol. 39 (21) , 7173-7193
Open AccessValue-correlation controlled integer-valued autoregressive process in random environment
The manuscript presents a new integer-valued first-order autoregressive process in a random environment, which is governed by two control processes. The first process defines the marginal distribu-tion, while the second regulates the correlation structure within the model. The properties of the proposed model are examined in detail, providing insights into its theoretical foundations and practical implications. Two methods for estimating the unknown parameters are introduced: the Yule-Walker estimator and the conditional maximum likelihood estimator. A series of simulations assesses the efficiency of these esti-mation techniques and demonstrates their performance across various scenarios. The effectiveness of the introduced model is further evaluated through its application to real-life data.
- Open Access
Morphological and Anatomical Differentiation of Potamogeton gramineus in Relation to the Presence of Invasive Species Elodea nuttallii: A Case Study from Vlasina Lake, Serbia
Elodea nuttallii represents non-native and highly invasive species in Europe that significantly influence freshwater plant communities by decreasing the diversity of native species. This study aimed to determine whether the morphological and anatomical features of Potamogeton gramineus, a native species in Vlasina Lake, differ between sites where it coexists with E. nuttallii and those where E. nuttallii is not present. Environmental variables such as water depth, temperature, pH, conductivity, saturation, and O2 concentration were included in the analysis. Analyses were conducted on 32 morphological and anatomical features of P. gramineus collected from six sites within Vlasina Lake, comprising three sites where E. nuttallii was present and three sites where it was absent. The datasets containing morphometric and environmental variables underwent analysis using standard univariate techniques (Descriptive, ANOVA), Tukey’s Honest Significant Difference (HSD) test, Student’s t-test, and the Mann–Whitney U test, as well as multivariate statistical methods such as Canonical Discriminant Analysis (CDA). The results show the presence of morphological differentiation among P. gramineus individuals across the analyzed sites. These findings suggest that morphological and anatomical features, such as epidermis, mesophyll, palisade, and aerenchyma tissue thickness in floating leaves, number, length, width, and the surface area of stomata, as well as the width of submersed leaves and stem aerenchyma tissue thickness, effectively differentiate individuals that coexist with E. nuttallii and individuals that growth without its presence. Moreover, they indicate that P. gramineus exhibits a notable ability to modify its morphological traits in response to invasion.
Open AccessMorphological and Anatomical Differentiation of Potamogeton gramineus in Relation to the Presence of Invasive Species Elodea nuttallii: A Case Study from Vlasina Lake, Serbia
Elodea nuttallii represents non-native and highly invasive species in Europe that significantly influence freshwater plant communities by decreasing the diversity of native species. This study aimed to determine whether the morphological and anatomical features of Potamogeton gramineus, a native species in Vlasina Lake, differ between sites where it coexists with E. nuttallii and those where E. nuttallii is not present. Environmental variables such as water depth, temperature, pH, conductivity, saturation, and O2 concentration were included in the analysis. Analyses were conducted on 32 morphological and anatomical features of P. gramineus collected from six sites within Vlasina Lake, comprising three sites where E. nuttallii was present and three sites where it was absent. The datasets containing morphometric and environmental variables underwent analysis using standard univariate techniques (Descriptive, ANOVA), Tukey’s Honest Significant Difference (HSD) test, Student’s t-test, and the Mann–Whitney U test, as well as multivariate statistical methods such as Canonical Discriminant Analysis (CDA). The results show the presence of morphological differentiation among P. gramineus individuals across the analyzed sites. These findings suggest that morphological and anatomical features, such as epidermis, mesophyll, palisade, and aerenchyma tissue thickness in floating leaves, number, length, width, and the surface area of stomata, as well as the width of submersed leaves and stem aerenchyma tissue thickness, effectively differentiate individuals that coexist with E. nuttallii and individuals that growth without its presence. Moreover, they indicate that P. gramineus exhibits a notable ability to modify its morphological traits in response to invasion.
- 20 February 2024
- Published by Springer Nature in Plant Biosystems
- Vol. 158 (3) , 380-393
Using morpho-anatomical characters of leaves and stems for distinguishing Bolboschoenus taxa: a case study of Serbian species
- 20 February 2024
- Published by Springer Nature in Plant Biosystems
- Vol. 158 (3) , 380-393
Using morpho-anatomical characters of leaves and stems for distinguishing Bolboschoenus taxa: a case study of Serbian species
- Open Access
- 1 January 2022
- Published by Centre for Evaluation in Education and Science (CEON/CEES) in Glasnik Antropoloskog drustva Srbije
- No. 55,p. 1-10
Analysis of seasonal movement in mountain centers of Serbia, with a special overview of Kopaonik
Depopulation of mountains has been an intense and lengthy process in Serbia. This paper will analyze the potential of Kopaonik mountain for the development of tourism, which can be one of the key factors for the demographic revitalization of the entire mountain. In tourism development so far, Kopaonik mountain stands out as one of the most important carriers of mountain tourism in Serbia. The limiting factor in the development so far is the pronounced seasonality of tourism promotion. The aim of the research in the paper will be to identify the factors that influence seasonality phenomena as the main limiting factor for achieving better economic effects.
- 1 January 2022
- Published by Centre for Evaluation in Education and Science (CEON/CEES) in Glasnik Antropoloskog drustva Srbije
- No. 55,p. 1-10
Open AccessAnalysis of seasonal movement in mountain centers of Serbia, with a special overview of Kopaonik
Depopulation of mountains has been an intense and lengthy process in Serbia. This paper will analyze the potential of Kopaonik mountain for the development of tourism, which can be one of the key factors for the demographic revitalization of the entire mountain. In tourism development so far, Kopaonik mountain stands out as one of the most important carriers of mountain tourism in Serbia. The limiting factor in the development so far is the pronounced seasonality of tourism promotion. The aim of the research in the paper will be to identify the factors that influence seasonality phenomena as the main limiting factor for achieving better economic effects.
- Open Access
- 1 January 2022
- Published by National Library of Serbia in Srpski arhiv za celokupno lekarstvo
- Vol. 150 (5-6) , 274-280
Morphometric analysis of somatotropic and folliculostellate cells of human anterior pituitary during ageing
Introduction/Objective. In this study, we have pointed out the immunohistomorphometric characteristics of somatotropic (GH) and folliculostellate (FS) cells of the male pituitary gland during ageing. Methods. On histological sections of the pituitary gland of 14 male cadavers of different ages, the GH and FS cells were immunohistochemically labeled with corresponding antibodies, monoclonal anti-GH antibody, and polyclonal anti-S100 antibody, respectively. Immunopositive GH- and FS-cells were further morphometrically analyzed using ImageJ software. Results. The obtained results of morphometric analysis showed that the surface area of GH cells increased significantly with age. In these cells, the nuclear?cytoplasmic ratio gradually decreased and became significantly higher after the age of 70 years. The volume density of GH cells has not changed during ageing, while in FS cells this parameter significantly increased in the cases older than 70 years. The nuclear?cytoplasmic ratio of GH cells is negatively correlated with the volume density of FS cells. Conclusion. Based on the obtained results, we concluded that hypertrophy of GH and FS cells occurs in men with ageing and that correlation between the morphometric parameters of these two cell types indicates their mutual interaction.
- 1 January 2022
- Published by National Library of Serbia in Srpski arhiv za celokupno lekarstvo
- Vol. 150 (5-6) , 274-280
Open AccessMorphometric analysis of somatotropic and folliculostellate cells of human anterior pituitary during ageing
Introduction/Objective. In this study, we have pointed out the immunohistomorphometric characteristics of somatotropic (GH) and folliculostellate (FS) cells of the male pituitary gland during ageing. Methods. On histological sections of the pituitary gland of 14 male cadavers of different ages, the GH and FS cells were immunohistochemically labeled with corresponding antibodies, monoclonal anti-GH antibody, and polyclonal anti-S100 antibody, respectively. Immunopositive GH- and FS-cells were further morphometrically analyzed using ImageJ software. Results. The obtained results of morphometric analysis showed that the surface area of GH cells increased significantly with age. In these cells, the nuclear?cytoplasmic ratio gradually decreased and became significantly higher after the age of 70 years. The volume density of GH cells has not changed during ageing, while in FS cells this parameter significantly increased in the cases older than 70 years. The nuclear?cytoplasmic ratio of GH cells is negatively correlated with the volume density of FS cells. Conclusion. Based on the obtained results, we concluded that hypertrophy of GH and FS cells occurs in men with ageing and that correlation between the morphometric parameters of these two cell types indicates their mutual interaction.
- 19 November 2021
- Published by Springer Nature
- p. 211-225
Springs of Southeastern Serbia with a Focus on the Vlasina Plateau: Different Types of Challenges for the Macroinvertebrate Community
- 19 November 2021
- Published by Springer Nature
- p. 211-225
Springs of Southeastern Serbia with a Focus on the Vlasina Plateau: Different Types of Challenges for the Macroinvertebrate Community
- 9 October 2021
- Published by John Benjamins Publishing Company
- p. 227-254
“You are the best!”
This study investigates relational practices in developing email conversations between interactants performing various institutional roles at a Norwegian university. Three variables were examined: conversational progression, institutional roles, and social distance. Conversational progression was found to exert an influence on increasing familiarity in closing sequences. The openings and closings in the conversations between faculty members tended to orient to familiarity to a greater extent than in most other institutional role dyads, while the widest variety and the highest frequency of relational moves outside the framing moves appeared in faculty – PhD fellow conversations. Social distance was identified as a driver of deference in both openings and closings, while its influence on the frequency of occurrence of other relational moves was limited.
- 9 October 2021
- Published by John Benjamins Publishing Company
- p. 227-254
“You are the best!”
This study investigates relational practices in developing email conversations between interactants performing various institutional roles at a Norwegian university. Three variables were examined: conversational progression, institutional roles, and social distance. Conversational progression was found to exert an influence on increasing familiarity in closing sequences. The openings and closings in the conversations between faculty members tended to orient to familiarity to a greater extent than in most other institutional role dyads, while the widest variety and the highest frequency of relational moves outside the framing moves appeared in faculty – PhD fellow conversations. Social distance was identified as a driver of deference in both openings and closings, while its influence on the frequency of occurrence of other relational moves was limited.
- Open Access
- 1 January 2021
- Published by National Library of Serbia in Filomat
- Vol. 35 (13) , 4469-4482
Identifying latent components of the TINAR(1) model
In this paper we give a solution for the problem of identifying and predicting latent components of the integer-valued time series with skewed Skellam marginal distribution. At the beginning, expressions for latent components identification and prediction are derived. These expressions give us a possibility of revealing the values of two hidden, immeasurable components which affect the integer-valued time series with skewed Skellam marginals. Yule-Walker estimators of the unknown parameters are obtained. Also, quality of identification and lag-one prediction is tested on simulated data. At the very end, the model validation is performed in application on the real-life data.
- 1 January 2021
- Published by National Library of Serbia in Filomat
- Vol. 35 (13) , 4469-4482
Open AccessIdentifying latent components of the TINAR(1) model
In this paper we give a solution for the problem of identifying and predicting latent components of the integer-valued time series with skewed Skellam marginal distribution. At the beginning, expressions for latent components identification and prediction are derived. These expressions give us a possibility of revealing the values of two hidden, immeasurable components which affect the integer-valued time series with skewed Skellam marginals. Yule-Walker estimators of the unknown parameters are obtained. Also, quality of identification and lag-one prediction is tested on simulated data. At the very end, the model validation is performed in application on the real-life data.
- Open Access
- 26 July 2017
- Published by Walter de Gruyter GmbH in Revista Romana de Medicina de Laborator
- Vol. 25 (3) , 237-244
Prediction of acute pancreatitis severity via the combined analysis of inflammatory biomarkers and coagulation parameters
Introduction. Timely assessment of severity of acute pancreatitis is needed to avoid severe systemic complications by making optimal therapeutic approach and correct prognosis of the disease. The aim of the study was to establish the role of several inflammatory biomarkers and coagulation parameters in prediction of AP severity, and also to propose a mathematical formula which allows their combined use for the same purpose. Material and Methods. The prospective study included 70 patients with AP. The patients were divided into groups: mild (group I), moderate (group II) and severe AP (group III). All patients were further classified into two groups: group A (mild AP) and group B (moderate and severe AP). Biochemical markers, inflammatory biomarkers and coagulation factors were tested in all patients. Results. Based on the results of Mann-Whitney,s test, it can be concluded that groups A and B are significant different from each other for CRP (p<0.05). Using the Wald’s stepwise forward method, a prediction model with CRP, PCT, D-dimer1, D-dimer3, fibrinogen1 and fibrinogen3 parameters as predictors of the severity of AP was obtained. The percentage of successful prediction of moderate or severe AP based on this model was 76.9%. The use of ROC analysis with the introduced linear combination from the logistic regression yielded equally good or even better results in the assessment of the severity of AP with the combined use of analyzed parameters. Conclusion. The combined analyses of biohumoral markers and coagulation parameters presented in the form a mathematical formula enabled a more accurate, rapid, rational and clinically available prediction of the severity of AP.
- 26 July 2017
- Published by Walter de Gruyter GmbH in Revista Romana de Medicina de Laborator
- Vol. 25 (3) , 237-244
Open AccessPrediction of acute pancreatitis severity via the combined analysis of inflammatory biomarkers and coagulation parameters
Introduction. Timely assessment of severity of acute pancreatitis is needed to avoid severe systemic complications by making optimal therapeutic approach and correct prognosis of the disease. The aim of the study was to establish the role of several inflammatory biomarkers and coagulation parameters in prediction of AP severity, and also to propose a mathematical formula which allows their combined use for the same purpose. Material and Methods. The prospective study included 70 patients with AP. The patients were divided into groups: mild (group I), moderate (group II) and severe AP (group III). All patients were further classified into two groups: group A (mild AP) and group B (moderate and severe AP). Biochemical markers, inflammatory biomarkers and coagulation factors were tested in all patients. Results. Based on the results of Mann-Whitney,s test, it can be concluded that groups A and B are significant different from each other for CRP (p<0.05). Using the Wald’s stepwise forward method, a prediction model with CRP, PCT, D-dimer1, D-dimer3, fibrinogen1 and fibrinogen3 parameters as predictors of the severity of AP was obtained. The percentage of successful prediction of moderate or severe AP based on this model was 76.9%. The use of ROC analysis with the introduced linear combination from the logistic regression yielded equally good or even better results in the assessment of the severity of AP with the combined use of analyzed parameters. Conclusion. The combined analyses of biohumoral markers and coagulation parameters presented in the form a mathematical formula enabled a more accurate, rapid, rational and clinically available prediction of the severity of AP.