https://rfos.fon.bg.ac.rs/handle/123456789/61 Wed, 15 Apr 2026 17:12:29 GMT 2026-04-15T17:12:29Z https://rfos.fon.bg.ac.rs/handle/123456789/118 Title: Okvir za evaluaciju multimodalnih biometrijskih sistema; Framework for evaluation of multimodal biometric systems Authors: Milenković, Ivan Contributors: Simić, Dejan; Starčević, Dušan; Nikolić, Boško Abstract: Primena biometrijskih tehnologija danas u ljudskom društvu postaje sve češća, gotovo da možemo konstatovati da je ona deo naše svakodnevnice. Prilikom implementacije biometrijske autentikacije, svaki sistem ima svoje zahteve i ograničenja, u zavisnosti od konkretnog scenarija u kojem se sistem koristi. Za odabir odgovarajućeg biometrijskog modaliteta, kao i algoritama za rad sa biometrijskim modalitetom, neophodno je sprovesti odgovarajuću evaluaciju performansi rada biometrijskog sistema. Ipak, ovu evaluaciju nije uvek lako sprovesti, kako za unimodalne, tako i za multimodalne biometrijske sisteme. Čak i kada su dostupne javne baze biometrijskih podataka za evaluaciju algoritama određenog biometrijskog modaliteta, potrebno je prilagoditi rad sistema protokolu testiranja koji konkretna baza definiše. U slučaju multimodalnog pristupa, evaluacija se dodatno komplikuje usled upotrebe različitih algoritama za fuziju informacija. Kako u dostupnoj relevantnoj literaturi nije pronađen detaljan prikaz modela evaluacije multimodalnih biometrijskih sistema, a radi prevazilaženja ovih teškoća, u okviru ovog doktorata definisan je objedinjeni model evaluacije multimodalnih biometrijskih sistema. Za definisanje ovog modela primenjena je MDA (Model Driven Architecture) paradigma. U okviru objedinjenog modela dat je metamodel evaluacije multimodalnih biometrijskih sistema, koji predstavlja svojevrsnu ontologiju pojmova značajnih za ovu oblast. Primenom ovog metamodela, moguće je kreirati modele evaluacije različitih biometrijskih sistema. Na osnovu modela evaluacije multimodalnih biometrijskih sistema kreiran je prototip okvira za evaluaciju multimodalnih biometrijskih sistema. Pomoću predloženog okvira moguća je evaluacija performansi multimodalnog biometrijskog sistema u različitim slučajevima korišćenja. Eksperimentalni rezultati evaluacije nad konkretnom bazom i algoritmima pokazuju da primena okvira skraćuje za četiri puta vreme potrebno za evaluaciju. Razvijena je i nova metoda za analitičko određivanje praga osetljivosti u skladu sa postavljenim parametrima željenog ponašanja sistema. Na kraju, na primeru alata koji je koristio neke od funkcionalnosti okvira, prikazano je kako primena okvira može učiniti efikasnijim proces obrazovanja inženjera u oblasti biometrije.; Application of biometric technologies in our contemporary human society is getting more frequent, so we can almost state that biometric technologies are part of our everyday life. When implementing biometric authentication, each system has specific requirements and constraints, which depend on the actual scenario in which the system is being used. In order to choose the adequate biometric modality, and also a fitting algorithm for the chosen modality, it is necessary to perform an evaluation of the biometric system performance. However, this evaluation is not always easy to conduct. This fact is true for both the unimodal and multimodal biometric systems. Even when open biometrics databases are available for evaluation, it is necessary to adapt system to work with testing protocol of the chosen open database. Moreover, if the biometric system uses multiple biometric modalities, evaluation gets even more complicated because of different available fusion algorithms. In order to overcome these difficulties, as there is not a detailed model of multimodal biometric systems available in relevant literature, this thesis presents a unified multimodal biometric systems evaluation model. Presented model is based on MDA (Model Driven Architecture) paradigm. A part of the unified multimodal biometric systems evaluation model is the metamodel of multimodal biometric system evaluation, which represents an ontology of terms used in this domain. Based on unified multimodal biometric systems evaluation model, a prototype framework for multimodal biometrics systems evaluation has been created. By using proposed framework it is possible to evaluate performance of multimodal biometric system in different use cases. Experimental evaluation results based on used database and algorithms show that the use of framework shortens time necessary for evaluation to a quarter of previously required time. Also, a new analytical method for biometric system threshold optimization, based on the predefined desired system behavior was developed. As final, a learning tool based on some of the framework functionalities is used to show how the use of framework can make the process of educating engineers in the field of biometrics more efficient. Fri, 01 Jan 2021 00:00:00 GMT https://rfos.fon.bg.ac.rs/handle/123456789/118 2021-01-01T00:00:00Z https://rfos.fon.bg.ac.rs/handle/123456789/117 Title: Poboljšanje efektivnosti daunsajzinga kroz povećanje proceduralne pravednosti metoda za određivanje potrebnog broja izvršilaca; Improving the effectiveness of downsizing by increasing the procedural fairness of required employee number determination methods Authors: Komazec, Stefan Contributors: Čudanov, Mladen; Jaško, Ondrej; Janićijević, Nebojša Abstract: Disertacija predstavlja rezultat višegodišnjeg naučno-istraživačkog i praktičnog rada. U svom prvom delu, disertacija obuhvata pregled dostupne literature iz oblasti daunsajzinga. Posebna pažnja je posvećena tačnom definisanju pojma daunsajzing i jasnom razlikovanju od drugih koncepata koji se vezuju za organizacione promene. Prikazano je različito shvatanje daunsajzinga među istraživačima, pre svega u pogledu obuhvata i pitanja da li se odnosi samo na organizacione promene koje se direktno tiču smanjenja broja zaposlenih, ili na organizacione promene koje imaju za cilj generalno smanjenje troškova i povećanje efikasnosti. Daunsajzing je u disertaciji posmatran holistički, kao koncept i organizaciona promena koja ima svoje uzroke i posledice. Mnogobrojni uzroci mogu dovesti do daunsajzinga. Pregled uzroka dat u disertaciji, bez obzira da li se radi o uzrocima iz okruženja organizacije ili o internim uzrocima, ukazuje na značaj teme, odnosno ukazuje na to da će daunsajzing uvek biti sastavni deo poslovanja. Posebno poglavlje u okviru pregleda relevantne literature je posvećeno samom procesu daunsajzinga, gde je daunsajzing analiziran kroz različite pristupe, ali i kroz konkretne strategije i taktike za sprovođenje daunsajzinga, na čemu je postavljena teorijska osnova kasnijeg istraživanja, koje je u središtu ciljeva i naučnih doprinosa disertacije. Iako literatura obiluje radovima i istraživanjima koja se odnose na koncepte organizacione pravednosti prilikom daunsajzinga, razmatrajući sve njene elemente, ipak postoje određeni istraživački jazovi. Utvrđeno je da nije bilo istraživanja koja su u svom fokusu imala način na koji je donošena odluka o potrebnom broju zaposlenih u organizaciji ili organizacionoj jedinici. Drugim rečima, istraživanja se nisu bavila odlukom koliko ljudi u daunsajzingu treba da ostane bez posla, već je u kontekstu sprovođenja daunsajzinga fokus bio na odluci ko treba da ostane bez posla. Na osnovu toga je postavljeno opšte istraživačko pitanje u disertaciji, gde je cilj utvrditi da li i način donošenja odluke o potrebnom broju zaposlenih, odnosno metode primenjene u tu svrhu imaju uticaj na stavove zaposlenih koji posle daunsajzinga nastavljaju da rade u organizaciji. U svrhu kompletiranja istraživačkog pitanja, sa posebnom pažnjom su obrađeni različiti ishodi koji se mogu očekivati od daunsajzinga, gde je fokus bio na individualnim ishodima, pre nego na organizacionim. Osnovno istraživačko pitanje je pretočeno u opštu hipotezu, od koje su daljom razradom dobijene tri posebne i devet pojedinačnih hipoteza. Za potrebe istraživanja razvijeni su konstrukti koji se odnose na karakteristike metode za utvrđivanje potrebnog broja zaposlenih: jasnoća, validnost i tačnost. Sa druge strane, iz literature su preuzeti koncepti motivacije, zadovoljstva organizacijom i namere za napuštanje organizacije. Devet pojedinačnih hipoteza pretpostavljaju uticaj između svih navedenih konstrukata i formiraju konceptualni model istraživanja. Sprovedeno je empirijsko istraživanje, kojim je obuhvaćeno 556 ispitanika iz različitih industrija, sa različitih hijerarhijskih nivoa, iz kompanija različitog porekla i sa različitom Poboljšanje efektivnosti daunsajzinga kroz povećanje proceduralne pravednosti metoda za određivanje potrebnog broja izvršilaca III vrstom delatnosti. Istraživačke hipoteze, i konceptualni model generalno, su testirane modeliranjem strukturalne jednačine, odnosno primenom SEM analize (eng. „Structural Equation Modeling“).; The dissertation is the result of many years of scientific research and practical work. In its first part, the dissertation includes a review of the available literature in the field of downsizing. Special attention is paid to the precise definition of the term downsizing and a clear distinction from other concepts related to organizational change. A different comprehension of downsizing among researchers is presented, considering its scope and the question whether it refers only to organizational changes directly related to staff reductions, or to organizational changes that have the aim to generally reduce costs and increase efficiency. In this dissertation downsizing is viewed holistically, as a concept and organizational change that has its causes and consequences. Numerous causes can lead to downsizing, and an overview of the causes given in the dissertation, regardless of whether they are causes from the organization's environment or internal causes, indicates the importance of the topic, i.e. indicates that downsizing will always be an integral part of business. Within the review of relevant literature a special chapter is dedicated to the essence of the downsizing process, where downsizing is analyzed through different approaches to downsizing, but also through specific strategies and tactics for conducting downsizing, on which the theoretical basis of later research is laid, representing the center of the goals and scientific contributions of the dissertation. Although the literature abounds with papers and research related to the concepts of organizational justice in downsizing, considering all its elements, there are still certain research gaps. It was determined that there was no research that focused on the way the decision on the required number of employees in the organization or organizational unit was made. In other words, the focus of the research was not on the decision on how many employees should lose their jobs through downsizing, but rather on the decision on who should lose the job. Based on that, the general research question was asked in the dissertation, with the aim to determine whether the way of making a decision on the required number of employees, i.e. the methods applied for those purposes, have the impact on the attitude of employees who remain in organization after downsizing. In order to complete the research question, special attention was paid to the various outcomes that can be expected from downsizing, where the focus was on individual outcomes, rather than on organizational ones. The basic research question was translated into a general hypothesis, from which three special and nine individual hypotheses were obtained by further elaboration. For the purposes of the research, constructs have been developed that refer to the characteristics of the method for determining the required number of employees: clarity, validity and accuracy. On the other hand, the concepts of motivation, satisfaction with the organization and intentions to leave the organization are taken from the literature. Nine individual hypotheses assume the influence between all the above mentioned constructs and form a conceptual research model. Fri, 01 Jan 2021 00:00:00 GMT https://rfos.fon.bg.ac.rs/handle/123456789/117 2021-01-01T00:00:00Z https://rfos.fon.bg.ac.rs/handle/123456789/116 Title: Model za upravljanje ranim fazama inovacionih projekata; Model for managing the early stages of innovation projects Authors: Milutinović, Radul Contributors: Stošić, Biljana; Levi-Jakšić, Maja; Petrović, Dejan; Makajić-Nikolić, Dragana; Bojović, Nebojša Abstract: Upravljanje ranim fazama predstavlja veoma značajnu oblast upravljanja inovacijama i inovacionim projektima. Ovo mišljenje је potvrđeno na osnovu velikog broja referentnih publikacija koje se mogu naći u navedenoj oblasti, ali i zastupljenošću identifikovane teme u praksi. Imajući u vidu da je upravljanje ranim fazama specificirano kao kritičan faktor uspeha inovacionih projekata, te da predstavlja jedan od najvećih izazova u ovoj oblasti, posebna pažnja se mora posvetiti unapređenju efikasnosti ovog dela projekta. U tom smislu, predmet istraživanja doktorske disertacije se odnosi na unapređenje upravljanja ranim fazama inovacionih projekata, kako bi se izvršila optimizacija pojedinih faktora i svojstava projekata, kao što je slaba strukturiranost, neizvesnost, rizik i dvosmislenost ciljeva. Osnovni cilj doktorske disertacije jeste razvoj modela za upravljanje ranim fazama i odabir parametara koji utiču na kreiranje efikasnijeg izlaza iz ovog dela projekta. Dok postojeći pristupi upravljanja, uglavnom, razmatraju pojedinačne delove ranih faza, cilj ovog pristupa je da integriše sve pojedinačne delove i da ih poveže u celinu, tako da izlaz iz jednog dela predstavlja ulaz u sledeći. Novorazvijeni pristup obuhvata algoritam za upravljanje idejama; Model 1 za definisanje tehničkih parametara koji ulaze u sastav koncepta proizvoda; Model 2 za određivanje vrednosti tehničkih parametara kojima se obezbeđuje definisani kvalitet uz minimalne troškove; i Model 3 za određivanje vrednosti tehničkih parametara kojima se maksimizira kvalitet u okviru ograničenog budžeta. Definisani modeli omogućavaju donosiocima odluka da na osnovu eksperimenata nad različitim scenarijima odluče koje rešenje je optimalno u datom trenutku. Proces izvođenja, kao i rezultati eksperimenata, dokumentovani su u cilju kreiranja benčmarka za buduća istraživanja.; Management of the early stages represents a very important field of innovation management and innovation projects management. This opinion is proven by the large number of referent publications that can be found in this field and also by the presence of identified topic in practice. Stated that the early stages management is specified as a critical success factor of innovation projects, thus representing one of the biggest challenges in this area, special attention must be paid to improving the efficiency of this part of the project. Accordingly, the subject of the doctoral dissertation is to improve the management of the early stages of innovation projects, in order to optimize certain project factors and features, such as poor structure, uncertainty, risk and ambiguity of goals. The main goal of the doctoral dissertation is to develop a model for managing early stages and the selection of parameters that affect the creation of a more efficient output from this part of the project. While existing approaches mainly consider individual parts of the early stages, the goal of this approach is to integrate all individual parts and to connect them in a way that output from one part represents the input to the next one. The newly developed approach includes an algorithm that represents one of the possible ways of managing ideas; Model 1 for defining the technical parameters that will be a part of the product concept; Model 2 for determining the values of technical parameters that provide defined quality at minimal cost; and Model 3 for determining the values of technical parameters that maximize quality within a limited budget. Defined models allow decision makers to decide which solution is optimal at a given time based on experiments on different scenarios. The execution process, as well as the results of the experiments, are documented aiming to create a benchmark for future research. Fri, 01 Jan 2021 00:00:00 GMT https://rfos.fon.bg.ac.rs/handle/123456789/116 2021-01-01T00:00:00Z https://rfos.fon.bg.ac.rs/handle/123456789/115 Title: Adaptivni fazi sistem za algoritamsko trgovanje : interpolativni Bulov pristup; Adaptive fuzzy system for algorithmic trading : interpolative Boolean approach Authors: Rakićević, Aleksandar Contributors: Petrović, Bratislav; Suknović, Milija; Martić, Milan; Žarkić-Joksimović, Nevenka; Marković, Aleksandar Abstract: Tema ovog rada je adaptivni fazi sistem za algoritamsko trgovanje. Sistem je razvijen korišćenjem interpolativnog Bulovog pristupa fazi modelovanju, analizi podataka i upravljanju. Predloženi pristup uključuje interpolativne logičke modele za fazi prepoznavanje cenovnih obrazaca na tržištu, logički DuPont metod za automatizovanu analizu profitabilnosti preduzeća, interpolativni fazi kontroler za upravljanje trgovanjem i genetski algoritam za obučavanje interpolativnog fazi kontrolera radi otkrivanja strategija. Interpolativni Bulov pristup, zasnovan na interpolativnoj Bulovoj algebri, prevazilazi problem nekonzistentnosti fazi logike. Konstruisani adaptivni fazi sistem može samostalno, iz podataka, da otkrije uspešne strategije, primeni ih za algoritamsko trgovanje i adaptira u slučaju pada njihovih performansi. Uspešnost sistema testirana je na podacima sa američkog tržišta akcija, međunarodnog deviznog tržišta i tržišta kriptovaluta.; The topic of this thesis is adaptive fuzzy system for algorithmic trading. The system is developed using interpolative Boolean approach for fuzzy modeling, data analysis and control. The proposed approach includes interpolative logical models for fuzzy recognition of price patterns in market data, logical DuPont method for automated analysis of company’s profitability, interpolative fuzzy controller for trading and a genetic algorithm for extracting trading strategies by training interpolative fuzzy controller. Interpolative Boolean approach, based on interpolative Boolean agebra, solves the problem of fuzzy logic’s inconsistency with Boolean axioms. The proposed system can independently discover successful trading strategies from data, apply them for algorithmic trading and adapt in the case of performance deterioration. The system was tested on historical data from US equity, foreign exchange market and cryptocurrency market. Wed, 01 Jan 2020 00:00:00 GMT https://rfos.fon.bg.ac.rs/handle/123456789/115 2020-01-01T00:00:00Z