Search

The main task of planning documents is to achieve maximal rationality in the use of space, spatial resources, and balanced territorial development. The preparation of plans is regulated by a legislative framework, which embraces multiple phases and stakeholders. In a perfect planning process, it would be expected that all the elements are well coordinated and brought to common understanding, but in reality, obstacles and challenges can occur in any of these steps, especially in the implementation phase. Although a plan is fully prepared to be practiced, its implementation might be omitted. Therefore, this article analyzes the full process of spatial and urban planning from the perspective of plan implementation. The methodology is based on a combination of different data collection methods (interviews, fieldwork, direct observation) with the analysis of plans and the particular mention of those plans that picture the implementation issues the most. Also, legislative acts and semi-annual and annual reports on the achievements of the local government budget are analyzed. The approach indicates that plan implementation depends the most on the willingness of the government to perform changes in the system and to peruse punitive policy comprehensively. Although it is about implementation at the local level, the success primarily depends on clear definitions given in the legislative acts, freedom of the local communities to make their own decisions, and financial decentralization, side by side with the regional and local circumstances, institutional technical and staff capacities, and application of participatory planning that involves actors from various sectors.

Maize is one of the most important crop species worldwide, but also extremely susceptible to the effects of increasingly higher temperatures and drought during the summer and its flowering and grain filling stage. Different strategies are being utilized to ensure a satisfying yield potential and quality even in the extremely unfavourable environmental conditions, which are the result of climate change. Some of them are cropping pattern changes and sowing alterations, including earlier sowing. Since this implies the exposure to suboptimal temperatures during early developmental stages, it leads to a demand for developing maize lines tolerant to low temperatures during these stages. This research focuses on the first phase of germination, imbibition. Maize tolerance to low temperatures is a complex trait that includes different mechanisms and strategies that all work together to ensure adaptation and survival, such as cell membrane changes, antioxidative system activation, etc. Additionally, the low temperature response of different maize inbreds varies substantially, so recognizing the different ways in which they respond to low temperatures during imbibition and other earlier stages of development is crucial. To accomplish this, seeds of two maize inbred lines of contrasting susceptibility to low temperatures were selected and exposed to control (20° C) and low temperature (8°C) conditions during the first 24h of imbition andthen further analyzed to assess their response. This included germination and tetrazolium assays; ascertaining membrane integrity by evaluating cell leakage and lipid peroxidation; determining the antioxidative capacity by assessing superoxide dismutase (SOD) and catalase (CAT) activity; and expression analysis of four genes included in the low temperature response (gibberellin insensitive dwarf 1 gibberellin receptor, gid1; fatty acid desaturases 2 and 6, fad2 and fad6; plastid-lipid-associated 2 protein, pap2). The results showed that, while t, Kukuruz je usev od izuzetne važnosti za poljoprivredu, koji se takođe odlikuje značajnom osetljivošću na visoke temperature i sušu tokom letnjih meseci, odnosno tokom faze cvetanja i nalivanja zrna. Jedna od strategija obezbeđivanja zadovoljavajuće produktivnosti i prinosa kukuruza, uprkos nepovoljnim uslovima životne sredine, jeste ranija setva. Ipak u tom slučaju su biljke kukuruza izložene suboptimalnim temperaturama u ranim fazama razvića, što stvara potrebu za razvojem linija tolerantnih na niske temperature u tim fazama, što je i fokus ovog istraživanja. Odgovor kukuruza na niske temperature je kompleksan i uključuje niz različitih mehanizama i strategija koje obezbeđuju opstanak, a takođe zavisi i od samog genotipa, faze razvića, itd. Iz tog razloga, odgovor kukuruza na niske temperature (8°C) je posmatran tokom faze imbibicije, kod dva genotipa, i to na nivou promena genske ekspresije (gid1; fad2; fad6; pap2), aktivnosti antioksidativnih enzima (SOD, CAT), integriteta ćelijskih membrana, stopa germinacije, kao i samog preživljavanja. Rezultati ukazuju da postoje statistički značajne razlike između genotipova kada je reč o stopi preživljavanja, integritetu ćelijskih membrana i ekspresiji gena gid1i fad6 (p<0,05). Ovo istraživanje ukazuje na moguće mehanizme aklimatizacije na niske temperature tokom faze imbibicije i ističe značaj odabira odgovarajućih linija za raniju setvu.

Present-day airline industry is quite a competitive field and crew scheduling represents one of the crucial problems due to significant impact on the airline's cost. The crew scheduling problem is based on the assignment of crew members to operate different tasks of route. The main goal of this paper is to provide an analysis and a solution to one of the biggest problems detected on a small airport in the Serbia - the problem of ground crew scheduling. The paper presents the main characteristics, goals and limitations of a real-life problem identified at this small airport. In order to solve the problem, we developed a dynamic discrete simulation model. The model is developed in a spreadsheet environment of Microsoft Excel. Some of the main limitations found in the development of the model are strong constraints and multiple goals. The model presented in the paper is designed as a useful management tool for smaller airports and is aimed at the improvement of operative processes.

Bakar (Cu) je esencijalan mikroelement za biljke ali ukoliko je prisutan u višku ima fitotoksično dejstvo izazivajući oksidativni stres. Silicijum(Si) ne spada u esencijalne ali je jedini element koji ispoljava protektivan efekat na biljke u uslovima stresa. Primena silicijumove kiseline kod biljaka tretiranih toksičnim koncentracijama bakra ublažava štetna dejstva i povećava toleranciju na prisustvo bakra u višku. Mehanizmi delovanja silicijuma kod biljaka krastavca su detaljno proučeni i obuhvataju: smanjenje akumulacije Cu u biljkama preko snižene ekspresije gena odgovornih za njegovo usvajanje, imobilizaciju Cu u ćelijskim zidovima, povećanje sinteze Cu-liganada: organskih kiselina (citrat, malat i akonitat) i aminokiselina (nikocijanamin i histidin) koji smanjuju koncentraciju slobodnih, reaktivnih jona Cu unutar ćelije. Primena Si je uslovila smanjenje ekspresije mikro RNK (miR398 i miR408) koje predstavljaju glavne regulatore ekspresije gena i proteina koji sadrže Cu kao kofaktor. Posledično, akumulacija Cu-proteina Cu/Zn superoksid-dismutaze i plastocijanina koji vezujući Cu doprinose skladištenju viška jona Cu u biljnim ćelijama, je izraženija kod biljaka gajenih sa silicijumom. Razumevanje uloge i mehanizama delovanja Si u biljkama će doprineti njegovoj širokoj praktičnoj primeni u cilju povećanja rezistencije i tolerancije biljaka na stres., Although copper (Cu) is an essential microelement for plants, if present in excess it has a phytotoxic effect causing oxidative stress. Silicon (Si) is not essential, but it is the only element with protective effect to plants under stress. Application of silicic acid to plants treated with toxic concentrations of copper alleviates the harmful effects of excess copper, which results in increase tolerance in these plants. The protective mechanisms of Si in cucumber plants have been studied in more details and include a decrease in accumulation of Cu by downregulation of genes responsible for Cu uptake in plants, immobilization of Cu in the root cell walls as well as higher biosynthesis of Cu-ligands: organic acids (citrate, malate and aconitate) and amino acids (nicotianamine and histidine) that decrease the concentration of free, reactive Cu ions within the cell. The application of Si caused lower expression of microRNAs (miR398 and miR408) which are the main regulators of the expression of genes and proteins containing Cu as a cofactor. Consequently, the accumulation of Cu-proteins: Cu/Zn superoxide dismutase and plastocyanin, which are buffering sinks for excess Cu ions in plant cells, is more pronounced in plants grown with silicon. Better understanding the Simediated mechanisms in plants exposed to stress will contribute to its wider agronomic application in order to increase plant resistance and tolerance to stress.

The project EpiDoc XML Encoding of Roman Inscriptions from Serbia: Digitization of Ancient Epigraphic Heritage carried out by the Institute for Balkan Studies of the Serbian Academy of Sciences and Arts and funded by the Ministry of Culture and Information of the Republic of Serbia focuses on research, training, digital documenting and digital edition of Roman inscriptions from the territory of Serbia (the Roman province of Upper Moesia and parts of Lower Pannonia, Dalmatia and Thrace). The project’s primary aim is to form a national e-archive of inscriptions that will be encoded in EpiDoc TEI-XML format. This collection of digital corpora will represent a completely new epigraphic edition that is furthermore born-digital and ready for interchange. The project is envisioned as a core for networking and intensifying the collaboration between research and cultural institutions. Its mission is to educate, inspire and instigate collaborative work and new concepts in epigraphic research.

The present paper offers an outline of important evidence about the territories and the administration of the peregrine communities in Upper Moesia. The study of this type of communities can contribute to better understanding of the administrative history of this Roman province and help set the framework for the further research on this topic; this is especially important for the province of Upper Moesia since our knowledge of its ancient past is quite fragmentary., Apstrakt: Kada se napravi poređenje sa nekim drugim provincijama Rimskog carstva, generalno se ne zna mnogo o provinciji Gornjoj Meziji, a naročito kada je u pitanju upravna istorija i etničke i demografske prilike u njoj. Razlog tome treba tražiti pre svega u nedostatku dovoljno rečitih svedočanstava u izvorima, kako knjiškim tako i dokumentarnim. U ovom radu pažnja je usmerena na peregrine zajednice ili civitates peregrinae ove provincije. U pitanju su je jedan tip provincijskih administrativnih jedinica u koje je bilo organizovano domorodačko stanovništvo po uspostavljanju rimske uprave i koje je većinski bilo u statusu peregrina. Ne ulazeći u sva, već samo u pojedina važna pitanja nastoji se da se odrede okviri u kojima bi se u daljim istraživanjima mogli sprovesti metodološki principi, utvrđeni na drugom reprezentativnijem materijalu istog tipa administrativnih zajednica; daje se pregled najvažnijih svedočanstava o broju, rasporedu i teritorijama zajednica, te o pojedinim upravnim pitanjima i pretresaju najvažniji dosad izneti stavovi o njima.

Predmet doktorske disertacije je razvoj modela ekološkog upravljanja na nacionalnom nivou, integrisanjem metoda za procenu rizika, čime bi se stvorila podrška u odlučivanju na nacionalnom nivou u kreiranju zakona, politika regulativa i nacionalnih strategija, ali i omogućilo unapređenje upravljanja ekološkim rizicima na nacionalnom nivou. Takođe model treba da omogući holistički pristup pri proceni ekoloških rizika, uzimajući u obzir, nekada zanemarivane faktore rizika kao što su percepcija ekoloških rizika od strane šire javnosti, kao i sklonost ka ekološkom riziku različitih činilaca društva, od pojedinaca, organizacija i drugih društvenih grupacija. Model ekološkog upravljanja na nacionalnom nivou koji integriše metode procene rizika (Model EUnNN-IMPR) predstavljen u ovoj disertaciji se zasniva na hibridnom pristupu odnosno na korišćenju kvalitativnih i kvantitativnih metoda procene rizika u rangiranju promenljivih ekoloških rizika i rizika po životnu sredinu, čime se daje podrška odlučivanju pri određivanju prioriteta ciljeva ekološkog upravljanja. Cilj primene ovog modela je ukazivanje na kritične oblasti ekološkog upravljanja na nacionalnom nivou te davanje pravaca za definisanje strateških prioriteta za rešavanje ekoloških problema i zaštitu životne sredine. Jedna od koristi korišćenja takvog modela se ogleda u rezultatima primene modela ekološkog upravljanja na nacionalnom nivou koji se odnose na poboljšanje kvaliteta stanja životne sredine i zdravlja ljudi jedne zemlje, usled određivanja prioriteta pri ostvarivanju ciljeva ekološkog upravljanja zasnovanih na rangiranim ekološkim rizicima i rizicima po životnu sredinu uz uključivanje javnosti ne samo u proces percepcije i ocene rizika, već i u proces donošenja ekoloških odluka, čime se unapređuje podrška za 6 uspostavljanje transparentne i efektivne nacionalne politike i strategija zaštite životne sredine zemlje., The subject of the doctoral dissertation is the development of a model of environmental management at the national level by integrating risk assessment methods, which would support decision-making at the national level in the creation of laws, policies and national strategies, as well as improve the management of environmental risks at the national level. The model should also enable a holistic approach to the environmental risk assessment, taking into account sometimes previously neglected risk factors such as the perception of environmental risks by the general public, as well as the propensity for environmental risk of different participants of society, from individuals, organizations to other social groups. The environmental management model at the national level that integrates risk assessment methods presented in this dissertation is based on a hybrid approach, that is, using qualitative and quantitative risk assessment methods in ranking variable environmental and ecological risks, thereby supporting decision-making in prioritizing environmental management objectives. The purpose of the implementation of this model is to identify critical areas of environmental management at the national level and to provide directions for defining strategic priorities for solving environmental problems and environmental protection. One of the benefits of using such a model is reflected in the results of the implementation of environmental management model at a national level that relate to improving the quality of the environment and human health of a country, by setting priorities in achieving environmental management objectives based on ranked 8 environmental and ecological risks and by involving the public, not only in the process of risk perception and assessment, but also in the environmental decision-making process, thereby enhancing support for the establishment of a transparent and effective national policy and strategies for environmental protection of a country.

Protection and promotion of geoheritage has been changing and improving in recent decades, in line with the growing research in this field. Some research papers specifically study geomorphological geoheritage and introduce a special term - geomorphosite. Karst geomorphosites present specific objects of geomorphological heritage. On Kucaj and Beljanica mountains there is the largest karst area in Serbia. The purpose of this paper is to select and evaluate the scientific value of karst geomorphosites of Kucaj and Beljanica. More than one hundred karst geomorphosites were analyzed. In order to achieve more adequate coverage and representation, they were grouped into 25 geomorphotypes in three major geomorphological components of the investigated area. The results of this analysis highlighted the most important natural potentials of this area for establishing a karst-based geopark. More than 20 geomorphotypes have a positive recommendation index, which provides a good basis for further geotourism analysis.

Investigation of a CBRN incident has to be conducted in a systematic, objective, and timely manner in order to collect, preserve and analyse physical evidence from the associated site to be admissible in a criminal or civil proceeding, as well as in prosecution of acts of terrorism and war crimes. The inherently hazardous character of a crime scene involving CBRN agents adds complexity to investigative actions and requires additional expertise, equipment and conditions. The specific investigation strategies are necessary, together with the adapted CBRN detection equipment and forensics methods and procedures. This paper deals with the management of a CBRN crime scene and the specifics of investigation process, including interfaces between CBRN safety and security, and emergency response. The necessary framework, the conduct of operations, and an integrated command structure are outlined. The role of the technical support organisations with adequate CBRN expertises is emphasised.

The unique and at the same time very complex nature of the wine requires a more thorough selection process compared to other consumer products. Wine consumers are faced with a number of brands, grape varieties, and regions of origin, diverse tastes and prices. Using discrete choice experiment this study explores the importance that consumers in Serbia attach to five attributes of the local wine brands. Respondents were asked to evaluate 10 choice tasks, each with three different wines, and to choose the one that they would like to have with friends or family. The sample results indicate the high importance of the wine brand, while the price proved to be the least important. However, preference-based segmentation identified three clusters that differ primarily in the type of wine they favor, but also whether they like or not sweet and sparkling wines. It turns out that the price is a moderately important attribute in all clusters.

We propose a ski injury risk assessment model which allows ski resorts to take targeted and preventive actions towards critical ski regions. Currently, ski resorts mostly measure ski injury risk based on the ratios constituted of the number of injuries and number of skier days. We argue that this measure can be improved by using ski lift transportation, a more fine-grained measure of risk than skier days. As compared to skier days, which provide a birds-eye view on the risk level of a ski resort, ski lift transportation allows for a spatial-temporal granularity of risk calculation. In this paper, we calculate risk as a measure of injury rate, severity of injuries, and exposure. The model is validated on the data from Kopaonik ski resort, Serbia, which was gathered during five consecutive seasons on more than 17 million ski lift transportation records of nearly 1.45 million skier-days with 1889 reported injuries. During the observed period, the capacity of ski lift transportation system in Kopaonik increased by 58%, and injury rate increased nearly two times, which is due to the emergence of new transportation patterns, as we show in the paper. These patterns heavily influence distribution of injury rates across the ski resort. The proposed model allows for more targeted strategies for injury risk management.

Behavioral finance literature shows that in addition to Markowitz's rate of return and risk, private investors consider various other stock features. This paper discusses the problem of determining investors' preferences for portfolio selection criteria, as well as the problem of optimal portfolio determination from the investors' point of view. The study primarily focuses on private investors who are interested in one-time investments rather than stock trading. We use a discrete choice analysis and hierarchical Bayes method to measure individual investors' preferences, and a logit model to determine individual shares of preferences. We treat the share of preferences as the share of certain stocks in an optimal portfolio. The proposed methodology is illustrated by the example of companies whose stocks are traded on the Belgrade Stock Exchange. We measure respondents' preferences for companies, preferences for return rates, riskiness of stocks, and dividend rates. The results of comparing the performance of the resulting portfolio with the efficient frontier obtained using Markowitz's portfolio theory indicate its high efficiency, thus validating the proposed approach.

This paper investigates the problem of sending packets through network on the interdomain level under condition that required Quality of Service (QoS) is achieved on the end-to-end (E2E) path. Process of sending and routing packets by one of the packet dispersion strategies is modelled using Coloured Petri Nets (CPN). The PN model was simulated to find and capture disjoint routes which ensure that realized values of network performance metrics meet the required ones on the E2E level. Using Data Envelopment Analysis (DEA), network performance is evaluated (packet delay, jitter, packet loss rate) to find which of the previously mentioned disjoint paths are more efficient than others. Based on DEA analysis results, Petri Net (PN) is expanded into stochastic PN in order to implement adaptive packet dispersion strategy. This strategy implies that paths with less probability of losing packets are more probable to be used in packet routing. Thanks to efficiency analysis, probability of selecting service classes by domain has been determined so that greater quality of VoIP service is achieved, which highly depends on offered network performance.

Poslednjih godina veliki broj istraživanja je usmeren ka predikciji saobraćajnih gužvi. Različite statističke metode nisu pokazale značajan doprinos u prediktivnim performansama predikcije gužvi. Stoga se danas sve češće koriste algoritmi mašinskog učenja u cilju postizanja zadovoljavajućih rezultata predikcije. U ovoj disertaciji, predstavljena je metolodogija za klasifikaciju gužvi na bazi novorazvijenog modela Gausovih uslovnih slučajnih polja za strkturnu binarnu predikciju (GCRFBC). Ista je uspešno implementirana na realne probleme predviđanja gužvi. Metodlogija može biti uspešno primenjena na klasifikacione probleme opisane neusmerenim grafovima koji se ne mogu efektivno rešiti standardnim uslovnim slučajnim poljima (CRF). Novorazvijeni model, korišćen u metodologiji, je zasnovan na standardnim Gausovim uslovnim slučajnim poljima za regresiju (GCRF) koja su proširena latentnim promenljivim što daje brojne prednosti istom. Zahvaljujući latentnoj strukturi, učenje i zaključivanje u modelu ne zahteva komplikovane numeričke procedure, već može biti rešeno analitički. Pored toga, postojanje latentne strukture omogućava da model bude otvoren ka daljim poboljšanjima. Tri različita algoritma su razvijena: GCRFBCb (GCRFBC – Bajesovski), GCRFBCb-fast (GCRFBC – Bajesovski sa aproksimacijom) i GCRFBCnb (GCRFBC – ne-Bajesovski). Proširena metoda lokalne varijacione aproksimacije sigmoidne funkcije korišćena je za rešavanje integrala po latentnim promenljivim u Bajesovskoj verziji GCRFBC modela. U slučaju ne-Bajesovskog GCRFBC modela u učenju i zaključivanju je korišćena latentna promenljiva sa maksimalnom vrednošću funkcije gustine verovatnoće. Zaključivanje u GCRFBCb modelu je rešeno korišćenjem Njutn-Kotesovim formulama za jednodimenzionalnu integraciju. Usled velikog broja varijacionih parametara, računski trošak učenja je veliki, stoga je razvijena brza verzija Bajesovskog GCRFBC modela. Performanse modela su evaluirane na sintetičkim i realnim podacima. Pokazano je da, In the recent years, research committed in the field of congestion prediction present one of the most popular area of interest. A variety of novel methods for congestion prediction based on unstructured statistical (machine) learning have become the standard for congestion prediction. However, in this dissertaion I argue that structured machine (statistical) learning algorithms can significantly improve congestion prediction performances. In this dissertation, a Gaussian conditional random field model for structured binary classification (GCRFBC) is proposed for solving problems of congestion prediction. The model is applicable to classification problems with undirected graphs, intractable for standard classification CRFs. The model representation of GCRFBC is extended by latent variables which yield some appealing properties. Thanks to the GCRF latent structure, the model becomes tractable, efficient and open to improvements previously applied to GCRF regression models. In addition, the model allows for reduction of noise, that might appear if structures were defined directly between discrete outputs. Three different forms of the algorithm are presented: GCRFBCb (GCRGBC - Bayesian), GCRFBCbfast (GCRGBC - Bayesian approximation) and GCRFBCnb (GCRFBC - non-Bayesian). The extended method of local variational approximation of sigmoid function is used for solving empirical Bayes in Bayesian GCRFBCb variant, whereas MAP value of latent variables is the basis for learning and inference in the GCRFBCnb variant. The inference in GCRFBCb is solved by Newton-Cotes formulas for one-dimensional integration. Due to large numbers of variational parameters the computational costs of learning is significant, so fast version of GCRFBCb model is derived (GCRFBCb-fast). Models are evaluated on synthetic data and real data. It was shown that models achieve better congestion prediction performance than unstructured predictors. Furthermore, computational and memory complexity is evaluate

Последњих година велики број истраживања је усмерен ка предикцији саобраћајних гужви. Различите статистичке методе нису показале значајан допринос у предиктивним перформансама предикције гужви. Стога се данас све чешће користе алгоритми машинског учења у циљу постизања задовољавајућих резултата предикције. У овој дисертацији, представљена је метолодогија за класификацију гужви на бази новоразвијеног модела Гаусових условних случајних поља за стрктурну бинарну предикцију (GCRFBC). Иста је успешно имплементирана на реалне проблеме предвиђања гужви. Методлогија може бити успешно примењена на класификационе проблеме описане неусмереним графовима који се не могу ефективно решити стандардним условним случајним пољима (CRF). Новоразвијени модел, коришћен у методологији, је заснован на стандардним Гаусовим условним случајним пољима за регресију (GCRF) која су проширена латентним променљивим што даје бројне предности истом. Захваљујући латентној структури, учење и закључивање у моделу не захтева компликоване нумеричке процедуре, већ може бити решено аналитички. Поред тога, постојање латентне структуре омогућава да модел буде отворен ка даљим побољшањима. Три различита алгоритма су развијена: GCRFBCb (GCRFBC – Бајесовски), GCRFBCb-fast (GCRFBC – Бајесовски са апроксимацијом) и GCRFBCnb (GCRFBC – не-Бајесовски). Проширена метода локалне варијационе апроксимације сигмоидне функције коришћена је за решавање интеграла по латентним променљивим у Бајесовској верзији GCRFBC модела. У случају не-Бајесовског GCRFBC модела у учењу и закључивању је коришћена латентна променљива са максималном вредношћу функције густине вероватноће. Закључивање у GCRFBCb моделу је решено коришћењем Њутн-Котесовим формулама за једнодимензионалну интеграцију. Услед великог броја варијационих параметара, рачунски трошак учења је велики, стога је развијена брза верзија Бајесовског GCRFBC модела. Перформансе модела су евалуиране на синтетичким и реалним подацима. Показано је да се применом методлогије оствару, In the recent years, research committed in the field of congestion prediction present one of the most popular area of interest. A variety of novel methods for congestion prediction based on unstructured statistical (machine) learning have become the standard for congestion prediction. However, in this dissertaion I argue that structured machine (statistical) learning algorithms can significantly improve congestion prediction performances. In this dissertation, a Gaussian conditional random field model for structured binary classification (GCRFBC) is proposed for solving problems of congestion prediction. The model is applicable to classification problems with undirected graphs, intractable for standard classification CRFs. The model representation of GCRFBC is extended by latent variables which yield some appealing properties. Thanks to the GCRF latent structure, the model becomes tractable, efficient and open to improvements previously applied to GCRF regression models. In addition, the model allows for reduction of noise, that might appear if structures were defined directly between discrete outputs. Three different forms of the algorithm are presented: GCRFBCb (GCRGBC - Bayesian), GCRFBCbfast (GCRGBC - Bayesian approximation) and GCRFBCnb (GCRFBC - non-Bayesian). The extended method of local variational approximation of sigmoid function is used for solving empirical Bayes in Bayesian GCRFBCb variant, whereas MAP value of latent variables is the basis for learning and inference in the GCRFBCnb variant. The inference in GCRFBCb is solved by Newton-Cotes formulas for one-dimensional integration. Due to large numbers of variational parameters the computational costs of learning is significant, so fast version of GCRFBCb model is derived (GCRFBCb-fast). Models are evaluated on synthetic data and real data. It was shown that models achieve better congestion prediction performance than unstructured predictors. Furthermore, computational and memory complexity is evaluate

Predmet istraživanja ove doktorske disertacije je razvoj novog pristupa selekciji konsultanata i njihovo optimalno raspoređivanje na određene pozicije (aktivnosti) na jednom ili više projekata. Adekvatan plan rasporeda konsultanata se prepoznaje kao jedan od najvažnijih faktora za efikasnu i pravovremenu realizaciju jednog ili više projekata čije aktivnosti mogu da se odvijaju istovremeno, a da se pri tome poštuju kadrovska i budžetska ograničenja. Ovaj problem je nepolinomijalne složenosti (NP težak) odnosno postaje sve složeniji sa povećanjem dimenzija odnosno broja projekata, broja kosultanata i aktivnosti koje je neophodno realizovati. Donosilac odluke bi trebalo da se vodi jedinstvenim kriterijumom pri određivanju rasporeda tako da uspešno kreira balansiran i efikasan plan rasporeda konsultanata. Efikasnost plana zavisi od različitih ulaznih i izlaznih kriterijuma kao što su obučenost i kvalifikacije konsultanta ali i njihova cena i ocene pri realizaciji sličnih aktivnosti u prošlosti. Očigledno je da je priroda kriterijuma različita i da oni po svojoj prirodi mogu biti i kvantitativni i kvalitativni, što dalje otežava modeliranje procesa i određivanje jedinstvene mere efikasnosti. Prema tome, osnovno je pitanje kako da se kreira efikasan raspored koji će obezbediti optimalno angažovanje raspoloživih konsultanata na više poslova na jednom ili više projekata koji se odvijaju istovremeno, posebno ukoliko nekoliko konsultanata može da realizuje istu aktivnost sa različitim učinkom odnosno nivoom perfomansi. U ovoj doktorskoj disertaciji se predlaže korišćenje i proširivanje modela analize obavijanja podataka (DEA - Data Envelopment Analysis) kao tehnike pomoću koje se integrišu svi ulazni i izlazni kriterijumi i kreira se jedna mera kojom se ocenjuje efikasnost. Predloženo proširenje se odnosi na kreiranje odgovarajućeg DEA modela mešovitog (celobrojnog) linearnog programiranja pomoću koga se istovremeno ocenjuje efikasnost i vrši raspoređivanje konsultanata na po, The main subject in this dissertation is development of a new approach to the selection of staff and their optimal assignment to certain positions in one or more simultaneous projects. An adequate staff assignment is recognized as one of the most important factors for the efficient and timely implementation of projects whose activities can take place at the same time. This problem is very complex and becomes more and more complicated with the increase in the number of projects, the number of staff and the activities that need to be realized. The decision maker should be guided by a unique criterion so that it successfully creates a balanced and efficient assignment plan. The efficiency depends on different input and output criteria, such as the training and qualifications of the staff, but also their costs and scores in the implementation of similar activities in the past. It is obvious that the nature of the criteria is different and they can be both quantitative and qualitative, which further complicates the process of interlacing and the determination of a single measure of efficiency. Therefore, the fundamental question is how to streamline an effective schedule that will ensure optimum engagement of available staff on multiple activities in one or more projects that take place at the same time, especially if several staff members can implement the same activity with different level of performance. This doctoral dissertation proposes using and modification of the Data Envelopment Analysis (DEA) as a technique that integrates all input and output criteria and creates a single measure of efficiency. The proposed modification refers to the creation of an appropriate mixed-integer linear programming DEA model, which simultaneously assesses the efficiency and make the assignment plan of the staff member to positions (activities) in one or more projects. As a result of optimization, a unique assessment of efficiency for each staff member for each potential activity is

The paper presents the application of a numerical method for the determination of the absorbed dose rate of gamma and X radiation in the dielectric thin layer of hafnium dioxide (RfO2) or SiO2, which is located in the structure of the MOS capacitor. Considering the radiation characteristics of the selected dielectrics, it can be concluded that there are advantages of RfO2 over SiO2 in the radiation field with high-energy X-rays. Similar radiation effects should be expected for the interaction of dielectric material with gamma ray photons originating from a Co-60 source, both in the dielectric layer with SiO2 and in that with RfO2. It can be concluded that for the same radiation absorbed dose into the MOS capacitor with RfO2, there are a greater number of generated electron-hole pairs, in which case the value of effective trapping efficiency is smaller than if SiO2 was used. © 2019 IEEE.

During the construction of nuclear facilities, radiotherapy centers, hospitals with radiological diagnostic facilities, and radioactive waste storage facilities, complex requirements for protection against ionizing radiation have been set up. In its basic principles, modern approaches in the design of radiation protection procedures were derived from empirical conclusions that were obtained by solving practical problems encountered by nuclear engineers and physicists. In the radiation protection projections, as well as in the formation of different technical and technological procedures in the radiation environment, one of the most important characteristics is the build-up factor of the material used. Based on experimental and theoretical research, empirical formulas for calculating the build-up factors are used to solve practical problems in radiation protection. With the development of information and computer technologies for calculating the build-up factor, other numerical methods are also used, of which a very attractive and efficient Monte Carlo calculation method is used. In this paper, the calculations with the empirical approximate formulas by Taylor and Berger, as well as the results obtained by using the software based on the Monte Carlo method, were used to determine the build-up of the factors in the protection of gamma and X radiation for several selected materials (lead, water, iron, etc.). By comparing the results obtained with the empirical approximations and Monte Carlo calculations, certain conclusions have been drawn that direct us to new critical considerations about the usefulness of using the chosen method of calculation for the build-up factor of materials found in modern technological processes using sources of gamma and X radiation.

Spreadsheets have been widely used in many business and scientific areas for modeling and simulation. Nevertheless, a number of research surveys have shown that these applications are particularly prone to errors. Some of the reasons for this error-proneness can be attributed to the fact that spreadsheet models are usually developed by the end-users and that standard software quality assurance processes are typically not applied. In addition, end-users often do not have a "clear picture" about the model structure and elements represented in the spreadsheet. Therefore, the aim of this paper is to present the error detection and debugging approach developed for dynamic discrete spreadsheet models of inventory control. Spreadsheet modeling of inventory dynamics is based on the utilization of the discrete time system control concept. This concept provides additional rules (common and specific constraints) used for error detection. Preliminary experiments show general applicability of the approach.

The beneficial effects of silicon (Si) have been shown on plants using reduction-based strategy for iron (Fe) acquisition. Here we investigated the influence of Si on Fe deficiency stress alleviation in barley (Hordeum vulgare), a crop plant which uses the chelation-based strategy for Fe acquisition. Analyses of chlorophyll content, ROS accumulation, antioxidative status, concentrations of Fe and other micronutrients, along with the expression of Strategy II genes were studied in response to Si supply. Si successfully ameliorated Fe deficiency in barley, diminishing chlorophyll and biomass loss, and improving the activity of antioxidative enzymes, resulting in lowered reactive oxidative species accumulation in the youngest leaves. Alleviation of Fe deficiency stress correlated well with the Si-induced increase of Fe content in the youngest leaves, while it was decreased in root. Moreover, Si nutrition lowered accumulation of other micronutrients in the youngest leaves of Fe deprived plants, by retaining them in the root. On the transcriptional level, Si led to an expedient increase in the expression of genes involved in Strategy II Fe acquisition in roots at the early stage of Fe deficiency stress, while decreasing their expression in a prolonged stress response. Expression of Strategy II genes was remarkably upregulated in the leaves of Si supplied plants. This study broadens the perspective of mechanisms of Si action, providing evidence for ameliorative effects of Si on Strategy II plants, including its influence on accumulation and distribution of microelements, as well as on the expression of the Strategy II genes.

Copper (Cu) toxicity in plants may lead to iron (Fe), zinc (Zn) and manganese (Mn) deficiencies. Here, we investigated the effect of Si and Fe supply on the concentrations of micronutrients and metal-chelating amino acids nicotianamine (NA) and histidine (His) in leaves of cucumber plants exposed to Cu in excess. Cucumber (Cucumis sativus L.) was treated with 10 mu M Cu, and additional 100 mu M Fe or/and 1.5 mM Si for five days. High Cu and decreased Zn, Fe and Mn concentrations were found in Cu treatment. Additional Fe supply had a more pronounced effect in decreasing Cu accumulation and improving the molar ratio between micronutrients as compared to the Si supply. However, the simultaneous supply of Fe and Si was the most effective treatment in alleviation of Cu-induced deficiency of Fe, Zn and Mn. Additional Fe supply increased the His but not NA concentration, while Si supply significantly increased both NA and His whereby the NA:Cu and His:Cu molar ratios exceeded the control values indicating that Si recruits Cu-chelation to achieve Cu tolerance. In conclusion, Si-mediated alleviation of Cu toxicity was directed toward Cu tolerance while Fe-alleviative effect was due to a dramatic decrease in Cu accumulation.

The paper presents the application of a numerical method for the determination of the absorbed dose rate of gamma and X radiation in the dielectric thin layer of hafnium dioxide (RfO2) or SiO2, which is located in the structure of the MOS capacitor. Considering the radiation characteristics of the selected dielectrics, it can be concluded that there are advantages of RfO2 over SiO2 in the radiation field with high-energy X-rays. Similar radiation effects should be expected for the interaction of dielectric material with gamma ray photons originating from a Co-60 source, both in the dielectric layer with SiO2 and in that with RfO2. It can be concluded that for the same radiation absorbed dose into the MOS capacitor with RfO2, there are a greater number of generated electron-hole pairs, in which case the value of effective trapping efficiency is smaller than if SiO2 was used. © 2019 IEEE.

Aluminium (Al) toxicity in acid soils is a global problem. Here, we investigated Al tolerance in high yielding winter wheat (Triticum aestivum L.) cultivars bred in Serbia. The common relative mot length (RRL) test for Al tolerance, and both physiological (malate efflux) and molecular (Aluminium-Activated Malate Transporter 1 [TaALMT1] expression) approaches were used for this characterization. Both moderately Al-tolerant cvs. Ljiljana and Arabeska showed significantly higher malate efflux rate from the root tips in comparison to moderately Al-sensitive cv. Pobeda and followed the RRL pattern. Irrespectively of Al supply, moderately Al-tolerant cultivars showed significantly higher relative TaALMT1 expression than the Al-sensitive ones. A considerably high level of Al tolerance was found in cv. Ljiljana, which showed the highest Al-induced malate efflux along with the highest constitutive expression level of TaALMT1 transcripts. Our results also demonstrate that Al tolerance is based on a constitutive trait of high TaALMT1 expression and malate efflux in wheat roots, resulting in a decrease in root length reduction.

У оквиру ове докторске дисертације истражена је подршка процесу оперативног планирања производње и управљања ресурсима у малим и средњим предузећима. Посебан фокус стављен је на активности распоређивања и терминирања, који чине најкомплексније планерске проблеме на оперативном нивоу. Мала и средња предузећа (МСП) представљају доминантну категорију предузећа у Републици Србији, а и шире у земљама Европске уније (ЕУ-28). МСП представљају најучесталију категорију и у производном сектору Републике Србије и ЕУ. Производња према поруџбини и жељама купаца и процесна организација производње представљају најзаступљенији облик производње у МСП, а Job Shop проблем распоређивања је саставни део процеса оперативног тј. краткорочног планирања производње у овој врсти предузећа. Имајући у виду да оперативно планирање непосредно претходи активностима трансформације предмета рада и стварању додатне вредности, посебан нагласак дат је на управљању производним ресурсима у процесу производње (сировине, материјали и други предмети рада, машине, алати и остала средства за рад, људски ресурси, време према коме се активности планирања реализују, новац тј. капитал, информације, тржиште). У дисертацији се даје преглед основних теоријских појмова: мала и средња предузећа, подршка МСП, процес управљања у МСП, планирање производње и управљање ресурсима. Поред тога, посебно је дефинисан процес оперативног планирања производње, са аспекта активности из којих се састоји. Распоређивање и терминирање у производњи су препознати као једни од најчешћих проблема из домена оперативног планирања производње. Они подразумевају распоређивање производа на машине, тј. проналажење редоследа радних задатака или послова који треба да се изврше на свакој од више машина. Поред тога обухватају и дефинисање временских термина за почетак и завршетак појединих производних активности. Циљ прављења адекватног распореда, тј. редоследа операција, произилази из ограничења капацитета производних ресурса (радници, машине и друг, This dissertation proposal explores the support for operational production planning in small and medium-sized enterprises (SME). Special emphasis is put on scheduling activities that represent the most complex planning problem at the operational level. SMEs are the predominant category of enterprises both in Serbia and in the European Union (EU-28). SMEs also present the most frequent type of enterprises in the manufacturing sector in Serbia and the EU-28 countries. Make-to- Order (MTO) production and production organization according to the process are the most common forms of production in small and medium-sized manufacturing enterprises, and the Job Shop scheduling problem is an integral part of operational, i.e. short-term production planning in this group of enterprises. Bearing in mind that operational production planning is directly and closely related to the transformation of raw materials and semi-manufactured goods, which creates and adds value, we emphasize the need to manage manufacturing resources adequately in the production process (materials, machines, tools, labour force, time, money - i.e. capital, information, market). The dissertation proposal presents the overview of basic theoretical concepts: small and medium-sized enterprises, support for SMEs, the process of management in SMEs, production planning, and resource management. Additionally, the process of operational production planning is defined from the aspect of activities it encompasses. Production scheduling problems are observed as one of the most frequent issues in the field of operational production planning. They include scheduling products to machines, i.e. finding the sequence of tasks and activities that need to be performed on each of available machines. Also, it includes determination of time slots for the beginning and completion of certain manufacturing activities. The aim of an adequate schedule, i.e. sequence of operations, arises from the capacity constraints of production re

Planiranje rasporeda odeljenja u zdravstvenim ustanovama ili rekonstrukcije postojećih je kompleksan zadatak na strateškom nivou odlučivanja. Konfuzni rasporedi odeljenja mogu doprineti anksioznosti pacijenata, povećati utrošeno vreme i pređeni put pacijenata. Razvijanje kvantitativnih pristupa u rešavanju ovog problema datira od kraja sedamdesetih godina dvadesetog veka. U ovom radu predstavljen je pristup planiranju rasporeda odeljenja u okviru zdravstvene ustanove zasnovan na metodi promenljivih okolina (Variable Neighborhood Search - VNS). Predloženi pristup je testiran na skupu poznatih instanci među kojima su i dve realne instance: Univerzitetska klinika Regensburg, Nemačka i Ahmed Maher bolnica u Kairu, Egipat. Predložena heuristika pronalazi optimalna rešenja na svim instancama manjih dimenzija, dok na instancama većih dimenzija pronalazi rešenja koja su relativno bliska najboljim poznatim rešenjima iz literature., Planning the layout of departments in hospitals or reconstructing existing ones is a complex task at the strategic decision-making level. Confusing layouts can contribute to patients' anxiety, increase the time spent in clinical procedures and total distance travelled by patients. The development of quantitative approaches in solving this problem dates back to the late seventies of the twentieth century. In this paper, we present an approach to the hospital layout based on a Variable Neighborhood Search (VNS). The proposed approach was tested on artificial and two real-world instances: University Clinic Regensburg, Germany and Ahmed Maher Hospital in Cairo, Egypt. Presented computational results show that the proposed VNS quickly reaches all optimal solutions for small-scale instances obtained by exact solver and provides quality solutions on large-scale instances that are relatively close to the best known solutions in literature.

This paper considers a stochastic programming problem with a number of random parameters in the set of constraints. The method used for solving the problem is the iterative optimization- simulation approach. It consists of two phases: optimization phase, which includes solving a deterministic counterpart of the original chance-constrained problem, and a simulation phase in which the original constraints are checked using Monte Carlo simulation. One iteration corresponds to one scenario. If the decision maker is dissatisfied with the results, a new scenario is generated in which the deterministic values of stochastic parameters are changed in the direction that will provide a more robust solution. The deterministic counterpart in the new scenario is formulated depending on the result of the previous iteration. To that end, different heuristics are considered. The main goal is to provide a good insight on the optimization problem under uncertainty by performing a relatively small number of iterations. The general approach and results of the proposed framework are illustrated on an example of advertisement placement.