Search  English (United States) Hrvatski (Hrvatska)

innovative promotional partnershipArtificial Intelligence towards EU Multilingualism

Technical co-sponsorship

 

Presented papers written in English and published in the Conference proceedings will be submitted for posting to IEEE Xplore.

Authors are kindly asked to prepare presentations lasting no more than 10 minutes.

Event program
Tuesday, 5/26/2015 9:00 AM - 1:15 PM,
Camelia 2, Grand hotel Adriatic, Opatija
COST Action IC1206  -  BIOMETRICS, SOFT-BIOMETRICS AND DE-IDENTIFICATION 
Invited Presentation 
E. Kindt (KU Leuven, Belgium)
The Legislator and Biometric Data Processing in the EU: Where are We? 
Lectures  
1.S. Ribarić (Faculty of Electrical Engineering and Computing, Zagreb, Croatia)
De-identification for Privacy Protection in Multimedia Content - COST Action IC1206  
2.R. Lyngby Kristensen (DTU Compute, Technical University of Denmark, Kgs. Lyngby, Denmark), Z. Tan (Department of Electronic Systems, Aalborg University, Aalborg, Denmark), Z. Ma, J. Guo (Beijing University of Posts and Telecommunications, Beijing, China)
Binary Pattern Flavored Feature Extractors for Facial Expression Recognition: An Overview 
This paper conducts a survey of modern binary pattern flavored feature extractors applied to the Facial Expression Recognition (FER) problem. In total, 26 different feature extractors are included, of which six are selected for in depth description. In addition, the paper unifies important FER terminology, describes open challenges, and provides recommendations to scientific evaluation of FER systems. Lastly, it studies the facial expression recognition accuracy and blur invariance of the Local Frequency Descriptor. The paper seeks to bring together disjointed studies, and the main contribution is to provide a solid overview for future research.
3.K. Brkić, T. Hrkać, Z. Kalafatić (University of Zagreb, Faculty of Electrical Engiineering and Computing, Zagreb, Croatia)
Detecting Humans in Videos by Combining Heterogeneous Detectors 
Human detection is an actively researched problem in computer vision, and it has driven the development of a number of detectors, both problem-specific and general purpose. We investigate the performance of human detection in video sequences when three commonly used detectors are combined: background subtraction, boosted cascade of Haar-like classifiers and histograms of oriented gradients. We establish a baseline for each of the detectors and analyze the performance gain obtained by combining all three. We employ the commonly used PETS09 dataset. HOG and background subtraction perform resonably well when used individually, while the boosted cascade of Haar-like classifiers performs poorly regardless of the training set used (full body, lower body only, face only etc.). Our results indicate that overall detection performance is significantly improved when combining HOG with background subtraction. Adding the boosted cascade of Haar-like classifiers does not result in an improvement, which is to be expected given its low baseline result. We conclude that combining HOG and background subtraction is an elegant solution for utilizing the advantages of video data to improve detection performance while maintaining a low processing cost.
4.T. Marušić, Ž. Marušić (Faculty of Science and Education, Mostar, Bosnia and Herzegovina), Ž. Šeremet (Faculty of Mechanical Engineering and Computing, Mostar, Bosnia and Herzegovina)
Identification of Authors of Documents Based on Offline Signature Recognition 
Handwritten signature is used in various applications on daily basis. Whether one signs a contract, work documents, petition, or wants to approve a check payment, one will use personal signature to do all those things. In this paper we use this daily based biometric characterstic for identification and classification of students' papers and various exam documents used at Faculy of Science and Education, University of Mostar. In this paper we used OpenCv as image processing tool as well as for global feature extraction. As regards to classification method, we used Support Vector Machine.
5.M. Horvat (Polytechnic of Zagreb, Department of Computer Science and Information Technology, Zagreb, Croatia), D. Kukolja (University of Zagreb, Faculty of Electrical Engineering and Computing, Department of Electric Machin, Zagreb, Croatia), D. Ivanec (University of Zagreb, Faculty of Humanities and Social Sciences, Department of Psychology, Zagreb, Croatia)
Comparing affective responses to standardized pictures and videos: A study report 
Multimedia documents such as text, images, sounds or videos elicit emotional responses of different polarity and intensity in exposed human subjects. These stimuli are stored in affective multimedia databases. The problem of emotion processing is an important issue in Human-Computer Interaction and different interdisciplinary studies particularly those related to psychology and neuroscience. Accurate prediction of users’ attention and emotion has many practical applications such as the development of affective computer interfaces, multifaceted search engines, video-on-demand, Internet communication and video games. To this regard we present results of a study with N=10 participants to investigate the capability of standardized affective multimedia databases in stimulation of emotion. Each participant was exposed to picture and video stimuli with previously determined semantics and emotion. During exposure participants’ physiological signals were recorded and estimated for emotion in an off-line analysis. Participants reported their emotion states after each exposure session. The a posteriori and a priori emotion values were statistically compared. The experiment showed, among other reported results, that carefully designed video sequences induce a stronger and more accurate emotional reaction than pictures. Individual participants’ differences greatly influence the intensity and polarity of experienced emotion.
6.M. Horvat (Polytechnic of Zagreb, Department of Computer Science and Information Technology, Zagreb, Croatia), D. Kukolja (University of Zagreb, Faculty of Electrical Engineering and Computing, Department of Electric Machin, Zagreb, Croatia), D. Ivanec (University of Zagreb, Faculty of Humanities and Social Sciences, Department of Psychology, Zagreb, Croatia)
Retrieval of multimedia stimuli with semantic and emotional cues: Suggestions from a controlled study 
The ability to efficiently search pictures with annotated semantics and emotion is an important problem for Human-Computer Interaction with considerable interdisciplinary significance. Accuracy and speed of the multimedia retrieval process depends on the chosen metadata annotation model. The quality of such multifaceted retrieval is opposed to the potential complexity of data setup procedures and development of multimedia annotations. Additionally, a recent study has shown that databases of emotionally annotated multimedia are still being predominately searched manually which highlights the need to study this retrieval modality. To this regard we present a study with N = 75 participants aimed to evaluate the influence of keywords and dimensional emotions in manual retrieval of pictures. The study showed that if the multimedia database is comparatively small emotional annotations are sufficient to achieve a fast retrieval despite comparatively lesser overall accuracy. In a larger dataset semantic annotations became necessary for efficient retrieval although they contributed to a slower beginning of the search process. The experiment was performed in a controlled environment with a team of psychology experts. The results were statistically consistent with validates measures of the participants’ perception speed.
7.M. Horvat (Polytechnic of Zagreb, Department of Computer Science and Information Technology, Zagreb, Croatia), D. Duvnjak (University of Zagreb, Faculty of Electrical Engineering and Computing, Department of Telecommunicati, Zagreb, Croatia), D. Jug (Polytechnic of Zagreb, Department of Computer Science and Information Technology, Zagreb, Croatia)
GWAT: The Geneva Affective Picture Database WordNet Annotation Tool 
The Geneva Affective Picture Database WordNet Annotation Tool (GWAT) is a user-friendly web application for manual annotation of pictures in Geneva Affective Picture Database (GAPED) with WordNet. The annotation tool has an intuitive interface which can be efficiently used with very little technical training. A single picture may be labeled with many synsets allowing experts to describe semantics with different levels of detail. Noun, verb, adjective and adverb synsets can be keyword-searched and attached to a specific GAPED picture with their unique identification numbers. Changes are saved automatically in the tool’s relational database. The attached synsets can be reviewed, changed or deleted later. Additionally, GAPED pictures may be browsed in the tool’s user interface using simple commands where previously attached WordNet synsets are displayed alongside the pictures. Stored annotations can be exported from the tool’s database to different data formats and used in 3rd party applications if needed. Since GAPED does not define keywords of individual pictures but only a general category of picture groups, GWAT represents a significant improvement towards development of comprehensive picture semantics. The tool was developed with open technologies WordNet API, Apache, PHP5 and MySQL. It is freely available for scientific and non-commercial use.
8.R. Dorofeev, A. Dorofeev, O. Kostowskaya (Irkutsk National Research Technical University, Institute of Cybernetics, Department of Computer Eng, Irkutsk, Russian Federation)
Evaluation of the Quality of Voice over IP Telephony Using the Decision Support System "Qualimetric Expertise" 
The paper considers the use of qualimetric expertise and ontological approach for evaluation of the quality of voice transmission over IP telephony in order to choose which of operators provides telecommunication services of the best quality. On the base of the suggested approach, an application the of the developed decision support system "Qualimetric Expertise" for automation of the process of voice transmission quality evaluation and identification of the possible causes of the quality loss is shown.

CIS ACCEPTED PAPERS
 
Papers
 
1.İ. Işıklı Esener (Bilecik Seyh Edebali University, Bilecik, Turkey), S. Ergin (Eskisehir Osmangazi University, Eskisehir, Turkey), T. Yüksel (Bilecik Seyh Edebali University, Bilecik, Turkey)
A New Ensemble of Features for Breast Cancer Diagnosis 
In this paper, an automatic Computer Aided Diagnosis (CAD) system is completely designed for breast cancer diagnosis and it is verified on a publicly available mammogram dataset constructed during Image Retrieval in Medical Applications (IRMA) project. This database comprises three different patch types indicating the health status of a person. These types are normal, benign cancer, and malignant cancer and they are labeled by the radiologists for the IRMA project. In the realization of CAD system, all mammogram patches are firstly pre-processed performing a histogram equalization followed by Non-Local Means (NLM) filtering. Then, the Local Configuration Pattern (LCP) algorithm is performed for feature extraction. Besides, some statistical and frequency-domain features are concatenated to LCP-based feature vectors. The obtained new feature ensemble is used with four well-known classifiers which are Fisher’s Linear Discriminant Analysis (FLDA), Support Vector Machines (SVM), Decision Tree, and k-Nearest Neighbors (k-NN). A maximum of 94.67% recognition accuracy is attained utilizing the new feature ensemble whereas 90.60% was found if only LCP-based feature vectors are used. This consequence obviously reveals that the new feature ensemble is more representative than an LCP-based feature vector by itself.
2.M. Meža (Faculty of Electrical Engineering, University of Ljubljnaa, Ljubljana, Slovenia)
Preliminary Study on Detection of Atrial Fibrillation during Oscillometric Blood Pressure Measurement 
The paper presents a prototype method for detection of cardiac arrhythmia from pneumatic cuff pressure course, obtained during oscillometric blood pressure measurement. Oscillometric blood pressure measurement is an indirect method for blood pressure measure where pneumatic cuff is placed on patient’s extremity, usually upper arm, inflated and then gradually deflated. During deflation phase the pressure oscillations in the cuff caused by squeezed blood vessels are analyzed to determine patient’s blood pressure. Since presence of arrhythmia can be detected through observation of heart rhythm, it is possible to detect arrhythmia during blood pressure measurement. We have captured cuff pressure courses – samples from 34 patients of which 6 had arrhythmia present. From captured pressure courses we determined a set of features which we used to train a classifier for discrimination of samples with present arrhythmia from samples where arrhythmia is not present.
3.Ž. Knok (Polytechnic of Medimurje, Čakovec, Croatia), Z. Avdagić, S. Omanović (Faculty of Electrical Engineering, Sarajevo, Bosnia and Herzegovina)
Hybride neuro-fuzzy expert system for assessing diabetes risk 
There is an increasing number of diabetic diseases in the population. Early diagnosis increases possibility of healing and decreases healing expenses. This paper is focused on modeling of an expert system for assessing diabetes risk using artificial intelligence methods. The model is hierarchical with neuro-fuzzy blocks and voting on the output. Model optimization (learning) is done using the data of real patients acquired in a one public health institution. The expert system is implemented in the Matlab/Simulink environment. Validation of the system shows high sensitivity of 100% which is important for early diagnoses. Specificity is lower – only 90% which means that some patients are sent to further diagnoses although they are healthy. Results indicate that this modeling approach is applicable on assessing diabetes risk. The medical domain knowledge and experience contained in the real data is successfully transferred in the solution model – the proposed expert system. Validation of the proposed expert system indicates that such a system can be used as an auxiliary expert for early diagnoses of diabetes and improve the quality of health systems with lower diagnostic expenses.
4.D. Domović, T. Rolich (Tekstilno-tehnološki fakultet, Zagreb, Croatia)
Solving Strip-Packing Problem Using Sequence Pair 
Strip-packing problem is a combinatorial optimization problem with a goal of finding optimal positions of cutting patterns that meet the following conditions: each pair of cutting patterns must not overlap and they must be placed within the boundary of rectangular material area. An optimal placement that utilizes material usage needs to be created. Strip-packing problem is a common textile industry problem where it is also known as automated marker making problem. In this paper a sequence pair abstract representation is used to obtain a non-overlapping placement. For that reason cutting patterns are approximated with a rectangular bounding box and placement of rectangles is obtained first. Non-approximated cutting pattern’s positions and material dimensions are then decoded using longest common subsequence (LCS) method. These methods are combined with simulated annealing algorithm to find an optimal placement of cutting patterns. Experiments were conducted on 4 benchmark datasets: Marques and Albano (24 cutting parts), Dagli (30 cutting parts), and Mao (20 cutting parts) with material utilization of 83.54%, 82.2%, 69.86%, and 71.26% respectively.
5.T. Manojlović, I. Štajduhar (University of Rijeka, Faculty of Engineering, Rijeka, Croatia)
Predicting Stock Market Trends Using Random Forests: A Sample of the Zagreb Stock Exchange 
Stock market prediction is considered to be a challenging task for both investors and researchers, due to its profitability and intricate complexity. Highly accurate stock market predictive models are very often the basis for the construction of algorithms used in automated trading. In this paper, 5-days-ahead and 10-days-ahead predictive models are built using the random forests algorithm. The models are built on the historical data of the CROBEX index and on a few companies listed at the Zagreb Stock Exchange from various sectors. Several technical indicators, popular in quantitative analysis of stock markets, are selected as model inputs. The proposed method is empirically evaluated using stratified 10-fold cross-validation, achieving an average classification accuracy of 76.5% for 5-days-ahead models and 80.8% for 10-days-ahead models.
Tuesday, 5/26/2015 3:00 PM - 7:00 PM,
Camelia 2, Grand hotel Adriatic, Opatija
Papers 
1.L. Begic Fazlic, Z. Avdagic, I. Besic (Elektrotehnički fakultet Sarajevo , Sarajevo, Bosnia and Herzegovina)
GA-ANFIS Expert System Prototype for Detection of Tar Content in the Manufacturing Process 
The purpose of this study is to present novel GA-ANFIS expert system prototype for tar detection in cigarettes during manufacturing process. The proposed system combines capabilities of Adaptive Neuro-Fuzzy Inference System (ANFIS) and Genetic Algorithm (GA).The data recorded for different type of cigarettes are collected by special control quality equipment in real conditions inside cigarette factory. GA-ANFIS system performs optimization in two steps. In the first step it generates six different ANFIS structures, and after that, we have second level of GA optimization using given ANFIS structures resulting in optimal fuzzy model structure. Modeling and validation of the GA-ANFIS system approach is performed in MATLAB environment using validation data set that was not used in the process of training. Our earlier research results based on two different approaches (ANFIS approach and High-performance liquid chromatography approach) is also introduced. Performances of these three approaches are compared and novel expert system prototype shows better result related to training, testing and validation errors. It also more precisely shows that low yield tar cigarettes contain similar levels of nicotine opposite to high yield tar cigarettes while benzene, toluene, and xylene (BTX) levels rise along with increasing tar yields.
2.A. Jović, K. Brkić, N. Bogunović (Sveučilište u Zagrebu, Fakultet elektrotehnike i računarstva, ZAGREB, Croatia)
A review of feature selection methods with applications 
Feature selection (FS) methods can be used in data pre-processing to achieve efficient data reduction. This is useful for finding accurate data models. Since exhaustive search for optimal feature subset is infeasible in most cases, many search strategies have been proposed in literature. The usual applications of FS are in classification, clustering, and regression tasks. This review considers most of the commonly used FS techniques. Particular emphasis is on the application aspects. In addition to standard filter, wrapper, and embedded methods, we also provide insight into FS for recent hybrid approaches and other advanced topics.
3.L. Bošnjak, S. Karakatič, V. Podgorelec (Faculty of Electrical Engineering and Computer Science, University of Maribor, Maribor, Slovenia)
Using similarity-based selection in evolutionary design of decision trees 
When evaluating the process of building classification decision trees, it is necessary to assess the performance of constructed trees, as well as the speed and efficiency of the algorithm. Top-down induction algorithms are relatively simple and can quickly generate good solutions, however their deterministic nature often prevents them from finding globally optimal solutions. On the other hand, the evolutionary approach to decision tree building has yielded promising results by exploring and exploiting the entire search space. However, the standard evolutionary method of building decision trees uses the fitness-based selection of two trees for crossover, which can lead to premature convergence to a local, often sub-optimal solution. In order to maintain the diversity of the population over the course of evolution, we propose a novel method of selection that takes into consideration the similarity of trees in the crossover process, to prevent inbreeding. Several different approaches to evaluate the similarity between trees were designed and implemented. The approaches of both similar and diverse tree crossover were compared to the standard induction algorithm on twenty different data sets to determine the impact of similarity on the effectiveness and efficiency of the genetic algorithm.
4.E. Cherkashin (Institute of System Dynamics and Control Theory at SB RAS, Irkutsk, Russian Federation), K. Paskal (National Research Irkutsk State Technical University, Irkutsk, Russian Federation), I. Bychkov (Institute of System Dynamics and Control Theory at SB RAS, Irkutsk, Russian Federation)
A Semantic Markup Technique Based on Ontology Polysystem 
The described investigation proposes an approach to overcome known problem of site developers' participation in the real development of semantic web content. Most of developers are interested in rather practical aspects of site functioning: publishing data, earning money. The site integration in a way of the semantic web is too hard to be used in a regular site maintenance. This results in the necessity of development of site content management software integrated by means of semantic web as knowledge acquisition systems, where developers and users have to play a role of an information source for a system-integrated decision-making engine inducing formal knowledge on the base of data analysis. A general idea of an approach to representation and induction of a semantic markup (logical layer) for description of the text content (Internet sites, legal documents, and so on) is being described. The logical layer is generated on the base of analysis of changes introduced by user. The changes of the text and the logical layer are analyzed. The variant of the interpretation (e.g., error correction of a value or a new statement definition) of the changes is determined by means of user interview. The interpretation also depends on the results of analysis of user behavior, e.g., patterns of transitions between various kinds of documents. The theoretical basis of the technique is to use of a polysystem representation of ontologies for the domain. The presentation is a hierarchically fibered structure of concepts and relations, which are mapped between fibers by means of interpretation relations. An automation of document preparation activities in a notary office has been chosen as a testing ground for the technologies under development. The documents that are originated and used there have an important common property. The documents contain information which is represented as structural and nonstructural data equally likely.
5.A. Hmelnov, E. Fereferov, R. Fedorov, E. Cherkashin, I. Bychkov (Institute of System Dynamics and Control Theory at SB RAS, Irkutsk, Russian Federation)
A Descriptive Specification Tools for Information System Design and Configuration 
Software development technologies evolved from concrete algorithm design and their programmatic implemen¬tation to modeling the whole information systems (IS). The modeling approaches allows developers to describe the system in terms of domain, software design, properties of implementation platforms, and standard ways of the IS development such as programming traditions in a concrete firm. In the case if a new IS is being constructed, having an already implemented database, Data Driven Engineering (DDE) is used. The model of this IS is constructed by importing database structure and adding a semantic layer description reflecting the database structure to the domain terms. We describe a DDE-approach based on declarative spec-ifications of existing database schema and properties of the target application, as well as interpretation of the description into IS components. The interpretation in various aspects allows developer to construct user interface for database modification, subsystem of SQL query construction in terms of domain, etc. An example result is presented.
6.D. Helic, F. Geigl (TU Graz, Graz, Austria)
Importance of Network Nodes for Navigation with Fractional Knowledge 
Due to sheer size of today's information and social networks people and algorithms are limited in their knowledge about those networks. Since this knowledge is important in many processes taking place on such network, e.g. search and navigation, an interesting question is how this limited knowledge should be constituted and structured. For example, when agents navigate an information network the question is what are the good or important nodes and links to know about for an efficient navigation. In this paper we apply machine learning techniques to learn an optimal division of network nodes into two classes: nodes that are important for navigation and nodes that are not important for navigation. We learn this optimal classification by minimizing the expected conditional loss (using 0-1 loss function) with respect to the globally shortest path between two arbitrary nodes. For learning we resort to MCMC methods such as simulated annealing and a simple approach for random assignment of nodes to classes. We perform a series of experiments by applying our optimization algorithm to synthetic networks with different community structures to shed more light into the relation between network structure, the node structural properties such as various centrality measures, and node classes produced by the algorithm.
7.M. Kunaver, Š. Dobravec, A. Košir (Faculty of Electrical Engineering, Ljubljana, Slovenia)
Using latent features to measure the diversity of recommendation lists 
One of the well known issues with content recommender system is that they tend to become over-specialized, which often has a negative influence on user experience. This can be solved by diversification of the recommendation list, a process that implements a tradeoff between accuracy and diversity of recommended items. Normally, item metadata is used in the diversity measure. In certain cases however, the item metadata may not be available thus a different approach to measure diversity is required. The aim of this preliminary study is to determine whether latent features can be used to measure the diversity of recommended items. In order to determine this we generated recommendation lists for 43 different users using the Ldos comoda dataset. We then evaluated the diversity of these lists using the standard intra-list diversity measure. In addition we calculated the diversity of each list by comparing the latent features (calculated using the matrix factorization approach) of each item on the list. The comparison of both value sets showed that they show similar characteristics which implies that latent feature space offers an alternative method of evaluating item diversity when no metadata is present.
8.L. Mercep (Cognizant Technology Solutions, Munich, Germany)
Efficient Profiling and Distributed Synchronization of Statistical User Models for Content-Targeting and Real-time Applications 
This work suggests an improvement of the efficiency of exact inference on Bayesian statistical knowledge representation of users or customers interacting with different information processing systems. The statistical user models are assumed to be distributed over multiple spatially separated systems and the models are assumed to be used for different real-time purposes, each of which incurs different quality-of-service requirements. A metric for the efficiency is composed out of three components: The amount of actual data being transferred between the systems using and the systems storing the user models, the computational complexity necessary to start acting on the basis of received user model and the computational complexity of synchronizing the distributed statistical user models. The knowledge is exchanged in the form which is optimized for querying i.e. for its usage during the agent’s lifetime, and not in the more explicit form used during the learning process or during the manual knowledge base design. Secondary knowledge structures such as junction trees, which can be constructed during the agent’s lifetime using statistical analysis of queries (through observation) and using the explicit metadata attached to the query (such as query priority or maximal latency allowed for answering the query in the case of strict real-time inference), can drastically reduce the computational complexity of inference. In some cases, meeting the real-time requirements on knowledge updating and parallel inference is not feasible at all without constructing the secondary knowledge structures. This is even more the case if exact inference is required, meaning that the entire knowledge base stored in or available to a particular agent has to be evaluated before answering a query, eliminating the possibility for heuristical approaches. This paper shows that the highest efficiency of information transfer and computational reduction of exact inference can be achieved only after identifying the hot-spots in the available variants of users’ secondary statistical models, using junction trees as a concrete example. These frequently used secondary structures represent the actual meaning or utility of a user’s statistical model, since they are indirectly generated by the quality-of-service requirements of the model’s consumers. Practical usage of the approach is suggested in the area of semi-autonomous or autonomous vehicles, in which the HMI and vehicle sensors can be used as user input and providing personalized HMI content and added security and safety features can provide benefit to both the customer and to the mobility service provider or vehicle manufacturer. The value added through content targeting and safety increase can be used to offset the increased costs of the intelligent vehicle and intelligent traffic systems.
9.D. Brodic, M. Jevtic (University of Belgrade, Technical Faculty in Bor, Bor, Serbia), Z. Milivojevic (College of Applied Technical Sciences, Nis, Serbia), V. Tasić (Mining and Metallurgy Institute, Bor, Serbia)
Text Skew Estimation Based on the Horizontal Entropy Calculation 
This paper describes the algorithm for text skew estimation based on the entropy. The method consists of the three steps. The first step calculates the horizontal projection profiles. Then, the entropy for the rough angles is calculated. At the end, the entropy calculation for the smooth angles is performed. The entropy is calculated by creating the cost function. The position where the cost function has an absolute minimum determines the text skew angle. The experiment shows that the method correctly estimates text skew. Besides, it is robust and time efficient.
10.V. Letic (BH Telecom, Sarajevo, Bosnia and Herzegovina), Z. Avdagic, D. Bošković (University of Sarajevo, Sarajevo, Bosnia and Herzegovina)
Optimization of Ring-Star Transmission Problem in Telecommunication Systems Based on Ant Colony Algorithms 
Solving complex combinatorial optimization problems using classical algorithms is not efficient related to resources and time. To overcome that problem, we can use optimal parameters selection based on Ant Colony Optimization (ACO) algorithms. In this paper, we present algorithm for solving telecommunication network using ACO for searching optimal ring star network topology. We analyzed ant's optimization ability based on shortest path between the nest and food location. In our research we used: Ant System, Elitist Ant System, Rank-Based Ant System, and MAX-MIN Ant System. The program is developed using GNU C++ to prove the algorithm theoretical convergence through simulation on variety of topologies regarding to node numbers. The algorithm was adapted to solve design of telecommunication network, which connects terminals to concentrators using point-to-point connections. The algorithm's output is a star topology showing connections of concentrators in a ring creating Digital Data Service. Algorithm uses seventeen parameters, with thirteen metrics to evaluate configurations. Program validation is done using three different network node configurations for all four ACO algorithms, only changing two control parameters: speed of pheromone evaporation and existence of local search. The best path was evaluated based on: total time, number of iterations, ring size, and value of topology.
11.I. Štajduhar, G. Mauša (University of Rijeka, Faculty of Engineering, Rijeka, Croatia)
Using String Similarity Metrics for Automated Grading of SQL Statements 
Manual grading of structured query language (SQL) statements after an exam can be tedious and time consuming for the teaching assistant. Additionally, it can also be subjective to her current state of mind and, thus, prone to errors. In this paper we propose an automated method for grading individual SQL statements. The method uses several common and simple string similarity metrics for comparing the student devised statements against the reference statements. These are then used, along with the manually assigned grades, for building the predictive logistic regression model. The proposed method was evaluated on a dataset consisting of 314 pairs of student-reference statements, along with the discretized average grade assigned by three independent evaluators. The model achieved the expected classification accuracy of 75% and F-measure 0,802 on a binary class (majority class 56%), thus exhibiting its potential for real-life application. The model can be used as is with the suggested calculated features and reported learnt parameters, or adapted to other examiners' evaluation criteria, presuming their willingness to build manually graded datasets of their own.
12.M. Schatten, I. Tomičić, B. Okreša Đurić (Laboratorij za umjetnu inteligenciju, Fakultet organizacije i informatike, Varaždin, Croatia)
Multi-agent Modeling Methods for Massivley Multi-Player On-Line Role-Playing Games 
Massively Multi-Player On-Line Role-Playing Games (MMORPG) give us the opportunity to study two important aspects of computing: (1) large-scale virtual social interaction of people (players) and (2) the design, development and coordination of large-scale distributed artificial inteligence (AI). A common denominator for both aspects are the methods used to study them: social interaction can be descibed and simulated using agent-based models (ABM) whilst distributed AI is commonly modelled in terms of multi-agent systems (MAS). The important question to ask in both perspectives is how do agents organize in order to perform their tasks and reach their objectives? In the following paper we will present and overview of agent-based approaches to modeling MMORPG's agents including human players, artificial players (bots) and non-playing characters (NPCs). We will put a special accent on useful agent behaviours, coordination and consensus mechanisms as well as organization and social network structure of agent societies.
13.B. Karahodza, D. Donko, H. Šupić (University in Sarajevo, Faculty of Electrical Engineering, Sarajevo, Bosnia and Herzegovina)
Temporal Dynamics of Changes in Group User's Preferences in Recommender Systems 
Using contextual information in recommender systems is a subject of continuous improvement of rating prediction accuracy. Among others, information on temporal rating dynamics contain valuable data that establish foundation for discovering changes in both individual and group user's preferences. Such changes can be caused by multiple factors such as changes of individual user interests, changes in item popularity or other hidden patterns or events. In this paper an improved user-based collaborative filtering algorithm is presented that utilizes changes of group user's preferences over time. We also investigate temporal dynamics of changes in user's preferences within different item categories and propose time weight function that improves prediction accuracy of recommender systems.
14.S. Dobravec (University of Ljubljana, Faculty of Electrical Engineering, Ljubljana, Slovenia)
Predicting sports results using latent features: a case study 
Predicting sports results is normally a challenging task, even more in case of a sport that shows a highly stochastic nature. In football, for example, numerous features are tracked and combined with expert knowledge, yielding various predicting algorithms. Our work however, is based on a case where there is no expert knowledge available and the only data comes from previous match results. We built a goal score prediction model that uses latent features obtained from matrix factorization process. We also added a Naive Bayes Classifier to be able to predict outcome of the match. The algorithm has been tested on results of the FIFA World Cup 2014. We also built a match result predictor based on the betting quotas. As these are derived from a complex algorithms that encompass also the expert knowledge, our algorithm can be used to estimate accuracy of an expert knowledge-based system. This case study shows that there is no significant difference between the two algorithms that we tested and that the latent features may provide a valid substitute for real features, when the later ones are not available.
15.L. Brajković (Faculty of Teacher Education, University of Zagreb, Zagreb, Croatia), W. Afrić (University North, Koprivnica, Croatia), V. Afrić (Faculty of Philosophy and Social Sciences, Zagreb, Croatia)
Role playing as a knowledge management tool 
Role playing, besides its use as a gaming media, can also be used as a powerful simulation tool. Due to its versatile nature and all the possibilities that it can offer us in building imaginary worlds and situations, we can also apply that technique to simulate real life and realistic behaviour. In project and knowledge management it is possible to use Role playing for optimization and building specific knowledge. Applying the simulation in advance, can greatly help us to recognize, prepare and solve most problems that may occur, reducing overall cost and time of any project.
16.Z. Balaž (Tehničko veleučilište , Zagreb, Croatia), I. Puhalović (Hrvatska agencija za civilno zrakoplovstvo, Zagreb, Croatia), G. Bjelić (Tehničko veleučilište Zagreb, Elektrotehnički odjel, Zagreb, Croatia)
Nadogradnja inteligentnog sustava aerodromskog održavanja serijskih strujnih krugova 
Sustav svjetlosne signalizacije na zračnim lukama u funkciji je sigurnosti zračnog prometa i za više razine opremljenosti zračne luke, (kategorije zračne luke CAT II-III, A, B, C), sustav može biti kreiran s više podsustava kojima je moguće racionalnije održavanje. Ocjene pogonske pouzdanosti rezultat su zadovoljavajućih parametara o stanju, koji se nadziru i arhiviraju u baze podataka te se o njima uči preko baza znanja algoritmima inteligentnog sustava. U radu će biti prezentirane baze podataka i rezultati obrade inteligentnim sustavom provedenih istraživanja na aerodromima i ispitivanjima u laboratoriju. Referentni parametri se odnose na stanje otpora izolacije aerodromskih primarnih kabela serijskih strujnih krugova, sustava svjetlosne signalizacije koji je predmet redovnih inspekcijskog nadzora i bitna komponenta u funkciji sigurnosti zračne plovidbe. Znanjem iz inteligentnog sustava utvrđena je mogućnost unapređenja procedura preventivnog i prediktivnog održavanja. Procedure i planovi održavanja, kao što je revitalizacija ili zamjena postojeće opreme, bazirane su prema utvrđenim parametrima za mjerenje kojih je korišten ispitni uređaj nove generacije. Posebna sofisticiranost ispitnog uređaja je da osim sveobuhvatnog ispitivanja prikupljanja i obrade, relevantne podatke može trenutno, proslijediti na zahtijevano mjesto, bez posredovanja računalom. Te podatke, posebnim protokolima moguće je na primjer u akcidentnim slučajevima, na zahtjev inspekcijskog nadzora dostavljati online na bilo koje mjesto ali i arhivirati u bazu podataka. Demo-model inteligentnog sustava koji je na raspolaganju studentima specijalističkog studija Elektrotehničkog odjela i polaznici programa edukacije za stjecanje licence aerodromskih specijalista AERO TVZ, na Tehničkom veleučilištu u Zagrebu nadograđuje se za potrebe „rada u oblacima“ s navedenim specifičnim mogućnostima.

Basic information:
Chairs:

Slobodan Ribarić (Croatia), Andrea Budin (Croatia)

International Program Committee Chairman:

Petar Biljanović (Croatia)

International Program Committee:

Alberto Abello Gamazo (Spain), Slavko Amon (Slovenia), Vesna Anđelić (Croatia), Michael E. Auer (Austria), Mirta Baranović (Croatia), Ladjel Bellatreche (France), Eugen Brenner (Austria), Andrea Budin (Croatia), Željko Butković (Croatia), Željka Car (Croatia), Matjaž Colnarič (Slovenia), Alfredo Cuzzocrea (Italy), Marina Čičin-Šain (Croatia), Marko Delimar (Croatia), Todd Eavis (Canada), Maurizio Ferrari (Italy), Bekim Fetaji (Macedonia), Tihana Galinac Grbac (Croatia), Paolo Garza (Italy), Liljana Gavrilovska (Macedonia), Matteo Golfarelli (Italy), Stjepan Golubić (Croatia), Francesco Gregoretti (Italy), Stjepan Groš (Croatia), Niko Guid (Slovenia), Yike Guo (United Kingdom), Jaak Henno (Estonia), Ladislav Hluchy (Slovakia), Vlasta Hudek (Croatia), Željko Hutinski (Croatia), Mile Ivanda (Croatia), Hannu Jaakkola (Finland), Leonardo Jelenković (Croatia), Dragan Jevtić (Croatia), Robert Jones (Switzerland), Peter Kacsuk (Hungary), Aneta Karaivanova (Bulgaria), Dragan Knežević (Croatia), Mladen Mauher (Croatia), Igor Mekjavic (Slovenia), Branko Mikac (Croatia), Veljko Milutinović (Serbia), Vladimir Mrvoš (Croatia), Jadranko F. Novak (Croatia), Jesus Pardillo (Spain), Nikola Pavešić (Slovenia), Vladimir Peršić (Croatia), Goran Radić (Croatia), Slobodan Ribarić (Croatia), Janez Rozman (Slovenia), Karolj Skala (Croatia), Ivanka Sluganović (Croatia), Vlado Sruk (Croatia), Uroš Stanič (Slovenia), Ninoslav Stojadinović (Serbia), Jadranka Šunde (Australia), Aleksandar Szabo (Croatia), Laszlo Szirmay-Kalos (Hungary), Davor Šarić (Croatia), Dina Šimunić (Croatia), Zoran Šimunić (Croatia), Dejan Škvorc (Croatia), Antonio Teixeira (Portugal), Edvard Tijan (Croatia), A Min Tjoa (Austria), Roman Trobec (Slovenia), Sergio Uran (Croatia), Tibor Vámos (Hungary), Mladen Varga (Croatia), Marijana Vidas-Bubanja (Serbia), Boris Vrdoljak (Croatia), Robert Wrembel (Poland), Damjan Zazula (Slovenia)

Registration / Fees:
REGISTRATION / FEES
Price in EUR
Before May 11, 2015
After May 11, 2015
Members of MIPRO and IEEE
180
200
Students (undergraduate and graduate), primary and secondary school teachers
100
110
Others
200
220

Contact:

Slobodan Ribarić
Faculty of Electrical Engineering and Computing
Unska 3
HR-10000 Zagreb, Croatia

Phone: +385 1 612 99 52
Fax: +385 1 612 96 53
E-mail: slobodan.ribaric@fer.hr

 
Andrea Budin
Ericsson Nikola Tesla Inc.
Krapinska 45
HR-10000 Zagreb, Croatia

Phone:+385 1 365 34 23
Fax: +385 1 365 3548
E-mail: andrea.budin@ericsson.com

Opatija - 170 years of tourism:

Opatija – the cradle of European and Croatian tourism, a favourite destination of the aristocracy, film and music stars, artists, writers and visitors from all over the world, who come here every year to enjoy the charm of this Adriatic town – this year celebrates its 170th anniversary as a tourist resort.

This is a tradition that provides certain obligations, but is also a guarantee of quality. The reputation of a top destination that stretches back seventeen decades is today reflected in the wide range of facilities and services on offer that all together make Opatija an attractive destination for all seasons.

Opatija owes its unique image to its ideal location on the spot where the wooded slopes of Mount Učka descend all the way down to the coast, providing perfect shade along the thirteen-kilometre-long Lungomare seafront promenade. Just as the Opatija area is a meeting point of the sea and the mountain, its visual impression is a blend of different styles, as this is a melting pot where magnificent Central European elegance, playful Mediterranean charm and the historically-rich medieval architecture of the small towns in the hinterland come together. 

In addition to the architecture that leaves a strong impression on every visitor, especially when the town is viewed from the sea, and its lush parks and gardens that have been Opatija's trademarks since its beginnings as a tourist resort, Opatija also has hotels and restaurants whose quality ranks alongside that of any other European destination. Opatija's gastronomic offer is based on a Mediterranean cuisine rich in fresh fish and seafood and locally grown seasonal ingredients, while the traditional recipes of this region reveal a wealth of flavours and can be sampled in the area's numerous taverns. 

Opatija entered the European stage in the mid-19th century as a health resort for the European nobility, and health tourism has remained one of the main segments of the town's tourism offer right up to the present day. However, top medical experts and a wide range of spa & wellness services are just one of the reasons for visiting this town located at the top of Kvarner Bay. Also known as "the town of festivals", Opatija boasts a number of events throughout the year. The theatrical performances and concerts that take place at the magnificent Open Air Theatre are particularly impressive.


For more details please look at www.opatija.hr/ and www.opatija-tourism.hr/.

Download
 
News about event
Currently there are no news
 
Patrons - random
HEP ZagrebSveučilište u ZagrebuSveučilište u RijeciFER ZagrebPomorski fakultet Rijeka