Search

Your search keyword '"Petrović, Bratislav"' showing total 86 results
Start Over Author "Petrović, Bratislav"
86 results on '"Petrović, Bratislav"'

1. IFS-IBA Logical Aggregation with Frank t-norms

Author

Milošević, Pavle, Dragović, Ivana, Zukanović, Milica, Poledica, Ana, Petrović, Bratislav, Goos, Gerhard, Founding Editor, Hartmanis, Juris, Founding Editor, Bertino, Elisa, Editorial Board Member, Gao, Wen, Editorial Board Member, Steffen, Bernhard, Editorial Board Member, Yung, Moti, Editorial Board Member, Massanet, Sebastia, editor, Montes, Susana, editor, Ruiz-Aguilera, Daniel, editor, and González-Hidalgo, Manuel, editor

Published
2023
Full Text
View/download PDF

3. Forecasting Sovereign Credit Ratings Using Differential Evolution and Logic Aggregation in IBA Framework

Author

Jelinek, Srđan, Milošević, Pavle, Rakićević, Aleksandar, Petrović, Bratislav, Kacprzyk, Janusz, Series Editor, Gomide, Fernando, Advisory Editor, Kaynak, Okyay, Advisory Editor, Liu, Derong, Advisory Editor, Pedrycz, Witold, Advisory Editor, Polycarpou, Marios M., Advisory Editor, Rudas, Imre J., Advisory Editor, Wang, Jun, Advisory Editor, Kahraman, Cengiz, editor, Cebi, Selcuk, editor, Cevik Onar, Sezi, editor, Oztaysi, Basar, editor, Tolga, A. Cagri, editor, and Sari, Irem Ucal, editor

Published
2022
Full Text
View/download PDF

18. Bacterial infections in patients with liver cirrhosis and ascites

Author

Marković-Živković Bojana, Bjelaković Goran, Nagorni Aleksandar, Benedeto-Stojanov Daniela, Petrović Bratislav, Radisavljević Mirjana, and Petrović Gordana

Subjects
liver cirrhosis, bacterial infections, spontaneous bacterial peritonitis, Medicine
Abstract

Introduction. Bacterial infections are common complications and the cause of death in patients with cirrhosis and ascites. There is no standard method for a rapid and low-cost diagnosis, and its prognosis is poor. Objective. The aim of this study was to determine the etiology and frequency of bacterial infections in patients with liver cirrhosis of different etiology, and the influence of bacterial infections on the prognosis in patients with liver cirrhosis and ascites. Methods. Sixty-four patients with cirrhosis and ascites were included in the study. The diagnosis of spontaneous bacterial peritonitis was established based on the diagnostic abdominal paracentesis and the results of biochemical, cytological and microbiologic analysis of ascitic fluid. The diagnosis of urinary infection and pneumonia were made according to the standard criteria. Results. Spontaneous bacterial peritonitis was diagnosed in 23 (35.9%) patients, urinary infections in 16 (25%) and pneumonia in 11 (17.2%). Gram positive and gram negative bacteria in spontaneous bacterial peritonitis were etiologically almost equally represented (52%; 48%). The most frequent causes were Escherichia coli and Staphylococcus aureus. In 81% of patients urinary infections were caused by gram negative bacteria (Escherichia coli in 44%). The most frequent cause of pneumonia was Streptococcus pneumoniae (46%). Conclusion. Spontaneous bacterial peritonitis, urinary infections and bronchopneumonia are the most frequent bacterial infections in patients with liver cirrhosis and ascites. A timely recognition of bacterial infections and the initiation of treatment have a positive effect on the prognosis of such patients.

Published
2014
Full Text
View/download PDF

28. Interpolative Boolean Networks

Author

Dobrić, Vladimir, primary, Milošević, Pavle, additional, Rakićević, Aleksandar, additional, Petrović, Bratislav, additional, and Poledica, Ana, additional

Published
2017
Full Text
View/download PDF

30. LOGIC-BASED AGGREGATION METHODS FOR RANKING STUDENT APPLICANTS.

Author

MILOŠEVIĆ, Pavle, NEŠIĆ, Ivan, POLEDICA, Ana, RADOJEVIĆ, Dragan, and PETROVIĆ, Bratislav

Subjects
RATING of students, BOOLEAN algebra
Abstract

In this paper, we present logic-based aggregation models used for ranking student applicants and we compare them with a number of existing aggregation methods, each more complex than the previous one. The proposed models aim to include depen- dencies in the data using Logical aggregation (LA). LA is a aggregation method based on interpolative Boolean algebra (IBA), a consistent multi-valued realization of Boolean algebra. This technique is used for a Boolean consistent aggregation of attributes that are logically dependent. The comparison is performed in the case of student applicants for master programs at the University of Belgrade. We have shown that LA has some advantages over other presented aggregation methods. The software realization of all applied aggregation methods is also provided. This paper may be of interest not only for student ranking, but also for similar problems of ranking people e.g. employees, team members, etc. [ABSTRACT FROM AUTHOR]

Published
2017
Full Text
View/download PDF

36. A Fuzzy Logic-Based System for Enhancing Scrum Method.

Author

Stupar, Mihailo, Milošević, Pavle, and Petrović, Bratislav

Subjects
FUZZY logic, AGILE software development, COMPUTER software development, NEW product development, QUANTIFIERS (Linguistics)
Abstract

In this paper, we propose a decision support system for enhancing scrum methodology based on fuzzy logic. Scrum is a very popular agile methodology within the software and product development. In the basic scrum, requirements that describe a certain task do not have a clear interpretation. Also, the traditional model does not take into account the experience of the developers nor the logical dependencies of input variables. Fuzzy inference is particularly useful for this purpose, because it incorporates logic in inference process and inputs are presented using linguistic quantifiers. The proposed system consists of three main components: a fuzzy inference system, an aggregation operator and a feedback function. The aggregation function is used to aggregate task predictions in a single value that uniquely represent a specific task, while a feedback is employed to adjust an input variable to improve system performance. Furthermore, the proposed system is simulated with randomly generated inputs in order to analyse its behaviour. The predictions of the system are more accurate and with smaller deviation in the final iterations. [ABSTRACT FROM AUTHOR]

Published
2017
Full Text
View/download PDF

38. Industrial Cladding of Stainless Steel to Low Carbon Steel by Explosion

Author

Petrović, Stojadin, Kosec, Ladislav, Mateša, Branko, Petrović, Vuk, and Petrović, Bratislav

Subjects
cladding, stainless steel to low carbon steel, explosion
Abstract

The base material ( ASTM A516 Gr.70, size: 2970x9800x1500 mm) is claded by explosion with cladding material( ASTM A 240 TP304L, same size, thickness 4 and 5 mm). Explosion cladded plates are designed for process industry pressure vessels production. In paper are except of the explosion(joining-welding)cladding mechanism basics, extensive research on metallurgical particularities presented .

Published
1995

43. INTESTINAL TYPE OF VILLOUS ADENOMA OF GALLBLADDER.

Author

Petrović, Bratislav, Petrović, Aleksandar, Živković, Vesna, Malina, Mitić, and Petrović, Vanja

Subjects
*GALLBLADDER cancer, *ADENOMA, *PRECANCEROUS conditions, *GALLSTONES, *CHOLECYSTECTOMY, *MORPHOGENESIS, *POLYPS, *METAPLASIA, *B cell lymphoma, *DIAGNOSIS
Abstract

Precancerous conditions in the gallbladder are still a matter of interest of numerous investigators, but reliable and definite attitude is not jet available. Particularly, many regenerative/reparative changes during inflammation and cholelithiasis in the gallbladder cannot be classified with certainty as lesions which precede malignant alteration. However, it is considered that metaplastic-dysplastic changes, which appear in mucosa and can lead to malignant neoplasia, are one of the ways of cancer morphogenesis. This way is considered as non-neoplastic way of gallbladder carcinoma development. The other way of cancer genesis goes through precancerous neoplasia, i.e. adenoma. The aim of this article was to, upon a polyp found in the gallbladder after cholecystectomy in a 70-year-old patient, determine one of the ways of gallbladder carcinoma morphogenesis. Cholecystectomy was done after ultrasound finding of the gallbladder polyp. Macroscopically and microscopically, it was the intestinal type of villous adenoma with the third degree epithelial dysplasia, which designates that this adenoma has reliable signs for precancerous lesion, and confirms the adenoma-carcinoma sequence theory, already characteristic for large intestine, the opinion the authors agree with. [ABSTRACT FROM AUTHOR]

Published
2010

44. Hereditary Hamartomatous Gastrointestinal Polyposis Syndrome.

Author

Živković, Vesna, Pejović, Svetlana, Nagorni, Aleksandar, Petrović, Bratislav, Petrović, Aleksandar, and Ilić, Ivan

Subjects
HAMARTOMA, POLYPS, PEUTZ-Jeghers syndrome, COWDEN syndrome, CANCER cells, MEDICAL genetics
Abstract

Copyright of Acta Facultatis Medicae Naissensis is the property of Nis University, Faculty of Medicine and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Published
2010

45. DECIDUALNA TRANSFORMACIJA U ADENOMIOZI U TOKU TRUDNOĐE KAO UZROK HISTEREKTOMIJIE.

Author

JOVANOVIĆ, Božidar, PETROVIĆ, Aleksandar, and PETROVIĆ, Bratislav

Subjects
*ENDOMETRIOSIS, *ENDOMETRIUM, *FEMALE reproductive organ diseases, *PELVIC diseases, *HYSTERECTOMY, *STERILIZATION of women
Abstract

INTRODUCTION: Incidence of adenomyosis is estimated on 20% in all of extracted uteruses. Clinically, it usually exists asymptomatically, but could be followed with dysmenorrhoea, menorrhagia and pelvic pain. One third of patients with adenomyosis are sterile, however in other two thirds, conception and pregnancy could have normal development, with delivery without complications. One of possible complications related to adenomyosis is rupture of uterus during delivery. CASE REPORT: A pregnant woman at the end of the 10th lunar month of pregnancy, showed minimal labour contractions, following amnion rupture and the delivery was terminated by section ceasar. Subtotal hysterectomy was performed because of concominant profuse bleeding. Myometrium was occupied with decidual cells in trabecular arrangement, among which elongated endometrial glands were placed. Surrounding muscle fascicles showed atrophic, apoptic, and occasionally necrotic changes. Myometrial stroma was edematous, and infiltrated with mesenchymal cells, as well as with adipose cells, which were extending toward serosa. DISCUSSION: The patient had been receiving treatment, for a long period, because of sterility without clear etiological factor. A small number of similar cases ends with complications. Here, it is evident that abundant decidual transformation of stromal cells in adenomiosis leads to atrophy and necrosis of muscle cells. The reduction of uterine muscle mass causes atony and threating rupture, through separation of muscle cells, and therefore the absence of their synchronized contractions. CONCLUSION: Atony of uterine muscle could be caused by decidual transformation in adenomyosal fields, atrophy of muscle fascicles, edema of the stroma, mesenchymal transformation, and fibrosis. This leads to a decrease in myometrial contractions, and prolonged postpartal bleeding. [ABSTRACT FROM AUTHOR]

Published
2009
Full Text
View/download PDF

46. POSTPARTALNA HISTEREKTOMIJA KAO POSLEDICA HRONIČNOG MIOMETRITISA.

Author

Jovanović, Božidar, Petrović, Aleksandar, and Petrović, Bratislav

Subjects
*MYOMETRIUM, *PATHOLOGY, *POSTPARTUM depression, *HEMORRHAGE, *HYSTERECTOMY
Abstract

Introduction As a diffuse chronic inflammation, myometritis is very rere and usually follows after postpartal placenta remains or postabortion infections, but it can be also associated with endometrial or ascendent infection. Chronic myometritis is often followed by profuse bleeding, though in most cases it cannot be recognized as it is asymptomatic. Histologically, that chronic process is characterized by the presence of fibriosis within the muscles and mononuclear cells (lymphoplasmocytic and histiocytic) infiltration. Case report A 24 old woman's second child was delivered per vias naturalis but the next day the profuse bleeding occurred which would not stop even after repeated curretages and suspecting a case of placenta accreta and uterus atony, subtotal hysterectomy was performed. Histologically, the disappearance of the regular arrangement of the smooth muscles and stroma could be seen with the devastation of myometrium due to the diffuse reduction of its smooth muscle bundles and cells, as well as their atrophy, necrobiosis and apoptosis with the minimal preservation of the muscle bundles and little cell groups of the myometrium, an abundant presence of the fibrocollagene and myxoid transformed connective tissue, group cells similar to the mesenchymal tissue and adipocytes. Discussion It was not possible to find this variant of the changes on the myometrium in the available literature. The present case is about the clinically unknown asymptomatic myometritis, possibly developed in the postpartal period of the previous pregnancy. It is our opinion that it is most probably an autoagressive process directed towards the smooth muscle cells of the myometrium, as shown by their reduction and inflammatory cells composition, which plays an important role in the immune reactions (lymphocytes, plasma cells, eosinophilis, histocytes). Conclusion A subtotal hysterectomy was performed on a woman, 24 years old, who gave birth to her second child and had profuse postpartal bleeding in sprite of repeated curettages. On the basis of this uterus atony, there is the clinically non-manifested chronic myometritis. The chronic inflammation resulted in a subtotal reduction of myometrium muscle mass, its replacement sclerosis, the multiplication of adipocytes, mesenchymal cells, histolcytes lymphomonocytes and dissection of muscle fascicles. [ABSTRACT FROM AUTHOR]

Published
2008
Full Text
View/download PDF

47. Систем за предвиђање недоласка путника на лет заснован на техникама рачунарске интелигенције

Author

Vojtek, Nikola, Petrović, Bratislav, Mijatović, Ivana, Milošević, Pavle, Dragović, Ivana, and Bojković, Nataša

Subjects
case based reasoning, interpolative Boolean algebra, no-show passengers forecasting, airline industry, закључивање на основу случаја, интерполативна Булова алгебра, предвиђање „no-show“ путника, авио индустрија
Abstract

Тема овог рада представља предлог система за предвиђање путника који се неће појавити на лету („no-show“), који се заснива на техникама рачунарске интелигенције. Предвиђање броја „no-show“ путника представља специфичан и уско формулисан проблем који је већ дужи низ година веома актуелан у авио индустрији како са теоријског, тако и са практичног аспекта. На основу очекиваног броја „no show“ путника, као и других фактора, авио компаније доносе одлуку о додатном броју места који ће бити доступан кроз резервациони систем. На овај начин, авио компаније могу остварити додатан профит, поготову када се ради о летовима који су попуњени у потпуности. Предложени систем за предвиђање “no-show“ путника се састоји од две компоненте. Прва компонента се односи на избор најпрецизнијег модела за предвиђање, а друга на примену и валидацију система. Модел за предвиђање се састоји из алгоритма који се заснива на техници закључивања на основу случаја и интерполативне Булове алгебре. Даље, модел комбинује предлог који је генерисан од стране алгоритма и предлог који препоручује експерт. На овај начин предложени систем обједињује и узима у обзир објективну и субјективну димензију приликом предвиђања. Сличност између летова се израчунава коришћењем традиционалних мера (Eуклидска и Mенхетн) и ИБА мере сличности. Такође, ИБА приступ употпуњује традиционални алгоритам технике закључивања на основу случаја кроз омогућавање логичке агрегације вредности, односно моделовањем постојећих нелинеарних зависности између података. Примена предложеног система је представљена коришћењем података о лету на релацији Београд - Амстердам, за период од годину дана. Добијени резултати показују да је неопходно укључити препоруку експерта у процес предвиђања, као и да сам алгоритам није довољан да би се добили довољно прецизни резултати. Такође, добијени резултати указују да су модели који су засновани на ИБА приступу и који комбинују резултате алгоритма и препоруку експерта, прецизнији од модела који користе традиционалне мере за израчунавање сличности. Сходно томе, потврђено је да логички приступ моделовању сличности представља перспективан правац примене у оквируviii технике закључивања на основу случаја. Са практичне стране, предложено решење је једноставно за разумевање у погледу функционисања, и може се доста једноставно имплементирати и прилагодити специфичностима и операцијама авио компаније. In this doctorial dissertation no-show passengers prediction system based on computational intelligence techniques is proposed. Predicting no-show passengers represents a specific and concisely formulated problem that actively persists for a longer period of time in the airline industry from both theoretical and practical perspective. Based on the expected number of no-show passengers, as well as some other factors, airlines are making decisions about how many additional seats will be allowed for overbooking through reservation system. This way, airlines could make additional profit, especially when it comes to the high demanding flights that are fully booked. Proposed prediction system for no-show passengers consists of two major components. First component considers selecting the best performing prediction model from the available model pool, and the second component is related to the model validation and application. Prediction model is based on the algorithm that combines case based reasoning technique and interpolative Boolean algebra (IBA) approach. Furthermore, model combines prediction recommendation generated by algorithm and recommendation provided from the domain expert. This way, the proposed system considers and takes into account both objective and subjective dimension. Similarity between flights is determined using traditional metrics (Euclidean and Manhattan) and IBA similarity measure. Also, IBA approach is enhancing the conventional CBR algorithm by enabling logical aggregation of values, i.e. capturing existing nonlinear dependencies in the data. The usage of the proposed system is illustrated in the numerical example regarding a single leg flight on the Belgrade-Amsterdam route and covers a one-year period. The obtained results show the necessity to include expert knowledge in prediction process, i.e. the CBR algorithm used alone is insufficient to produce results that are accurate enough. Furthermore, the results are indicating that the IBA-based models that combine the results of the CBR algorithm and expert recommendations perform better than distance-based models. Therefore, it is confirmed that the logicbased approach of similarity modelling is the prospective direction within the CBR algorithm. From a practical side, proposed solution is easy for understanding from thex functional aspect, and could be easily implemented and adjusted according to airline operations.

Published
2021

48. Адаптивни фази систем за алгоритамско трговање : интерполативни Булов приступ

Author

Rakićević, Aleksandar, Petrović, Bratislav, Suknović, Milija, Martić, Milan, Žarkić-Joksimović, Nevenka, and Marković, Aleksandar

Subjects
adaptive fuzzy system, algorithmic trading, trading system, trading strategy, fuzzy logic, interpolative Boolean algebra, interpolative fuzzy controller, logical DuPont method, interpolative logical models, адаптивни фази систем, систем за алгоритамско трговање, стратегија за трговање, фази логика, интерполативна Булова алгебра, интерполативни фази контролер, логички ДуПонт метод, интерполативни логички модели
Abstract

Тема овог рада je адаптивни фази систем за алгоритамско трговање. Систем је развијен коришћењем интерполативног Буловог приступа фази моделовању, анализи података и управљању. Предложени приступ укључује интерполативне логичке моделе за фази препознавање ценовних образаца на тржишту, логички ДуПонт метод за аутоматизовану анализу профитабилности предузећа, интерполативни фази контролер за управљање трговањем и генетски алгоритам за обучавање интерполативног фази контролера ради откривања стратегија. Интерполативни Булов приступ, заснован на интерполативној Буловој алгебри, превазилази проблем неконзистентности фази логике. Конструисани адаптивни фази систем може самостално, из података, да открије успешне стратегије, примени их за алгоритамско трговање и адаптира у случају пада њихових перформанси. Успешност система тестирана је на подацима са америчког тржишта акција, међународног девизног тржишта и тржишта криптовалута. 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.

Published
2020

49. Моделовање концепата конзистентном фази логиком

Author

Dobrić, Vladimir, Petrović, Bratislav, Martić, Milan, and Radojević, Dragan

Subjects
concepts, categories, градација, композиционалност, gradation, категорије, Boolean networks, сложени системи, фази когнитивне мапе, fuzzy cognitive maps, концепти, consistent fuzzy logic, Булове мреже, compositionality, complex systems, конзистентна фази логика
Abstract

Концепти су менталне репрезентације категорија. Теорија концепата треба да објасни како концепти функционишу као категорије и како се компонују у сложене концепте и мисли. Композиционалност објашњава креативност, како коначни ум може имати бесконачан когнитивни капацитет. Кључно питање у теорији је да ли је степен примене концепта бинаран или градиран. Теорија градације додатно објашњава репрезентативност инстанци категорија, али је теоретичари класичног погледа критикују да не може објаснити композиционалност те закључују да је погрешна. У дисертацији су помоћу конзистентне фази логике формализовани и објашњени концепти у случају градације. Прототип теорија концепата формализована је у стандардном и случају логичких интеракција између атрибута. Теорија егземплара формализована је на два начина. Теорија граница категорија формализована је у биполарном оквиру. Каузална теорија концепата формализована је у стандардном, случају каузалних повратних спрега између атрибута и/ или логичких интеракција између узрока. Математички је објашњена композиционалност Булових концепата у складу са прототип, егземплар и теоријом граница категорија, под претпоставком да концепти имају векторску природу. Одговорено је на критику прототип теорије и оповргнута су два аргумента којима се она поткрепљује, да закони искључења трећег и неконтрадикције нису загарантовани у прототип оквиру и да неодређено много Булових концепата нема прототип иако га све компоненте имају. Моделовани су конкретни Булови концепти за које теоретичари класичног погледа сматрају да не могу бити објашњени у прототип оквиру. У наставку дисертације генерализована су два експертска формализма за сложенe системе, Булове мреже и фази когнитивне мапе. Дескриптивна снага формализама драстично је увећана чиме они постају употребљиви у знатно ширем спектру ситуација... Concepts are mental representations of categories. Theory of concepts has to explain how concepts function as categories and how they compose into complex concepts and thoughts. The compositionality explains creativity – how a finite mind can have an infinite cognitive capacity. Fundamental question in theory of concepts is whether a degree of conceptual application is binary or graded. Theory of gradation additionally explains representativeness of category members, but it is criticized by classical theoreticians that it cannot explain compositionality, and thus, it is considered wrong. In this doctoral dissertation, concepts in the case of gradation are formalized and explained using consistent fuzzy logic. Prototype theory of concepts is formalized in standard case and in case of logical interactions between attributes. Exemplar theory of concepts is formalized in two ways. Boundary theory of concepts is formalized in bipolar framework. Causal theory of concepts is formalized in standard case, in case of causal feedback between attributes and/ or logical interactions between causes. The compositionality of Boolean concepts is mathematically explained according to the prototype, exemplar and boundary theory, under the assumption that concepts have vector nature. The critique of the prototype theory has been addressed, and two arguments that support the critique are refuted - that laws of excluded middle and noncontradiction are not secured in the prototype framework, and that there isn’t any prototype for indefinitely many Boolean concepts even though all components have one. Specific Boolean concepts, which are considered unexplainable in the prototype framework by classical theoreticians, are modelled. In the rest of the dissertation, two expert-based formalisms for complex systems, Boolean networks and fuzzy cognitive maps, are generalized. The descriptive power of two formalisms is drastically increased, which allows their usage in much broader spectrum of cases...

Published
2017

50. Логички приступ моделовању сличности

Author

Poledica, Ana, Petrović, Bratislav, Martić, Milan, and Radojević, Dragan

Subjects
logic-based similarity measures, classification, класификација, логичке мере сличности, класификација заснована на сличности, моделовање сличности, интерполативна Булова алгебра, similarity modeling, interpolative Boolean algebra, параметарске мере сличности, parameterized similarity measures, similarity-based classification
Abstract

У овој докторској дисертацији уведен је логички приступ моделовању сличности који је заснован на интерполативној Буловој алгебри. За мерење сличности, предложене су нове интерпретабилне логичке мере, параметарске и непараметарске, као и дескриптивни оператор агегације – логичка агрегација. Поред пружања теоријске основе, у овом истраживању посебна пажња је посвећена емпиријској анализи. У сврху валидације дефинисаних мера уведена је логичка класификација заснована на ИБА сличности. За све уведене мере извршена је евалуација и поређење на реалним подацима из домена медицине, где је показано да увођење параметара унапређује резултате класификације. На крају су приказане могућности за конструисање логичких класификатора заснованих на експертским функцијама сличности на проблему предвиђања банкротства предузећа. In this doctoral thesis, a logical approach to similarity modeling based on interpolative Boolean algebra is introduced. Novel interpretable logical measures, both nonparametric and parametrized, are introduced for measuring similarity together with logical aggregation as a descriptive aggregation operator. Besides the theоretical background, in this research special attention is devoted to empirical analysis. For validation purposes, logical classification based on IBA similarity is introduced. Defined logical measures are evaluated and compared in the case of medical data, and it is shown that parameterized measures improve classification results. Finally, the benefits of logic-based classifiers with expert similarity functions are presented on the problem of corporate bankruptcy prediction.

Published
2016

Catalog

Books, media, physical & digital resources
See catalog results