ORODJA ZA STROJNO UČENJE - KLASIFIKACIJAAndreja Zacirkovnik
, 2013, diplomsko delo
Opis: Diplomsko delo vsebuje teoretično osnovo strojnega učenja, podrobnejši opis podatkovnega rudarjenja ter njegove metode, natančneje klasifikacije. V drugem, praktičnem delu smo opisali tri prosto dostopna orodja, ki podpirajo strojno učenje ter metodo klasifikacije. Orodja smo testirali ter izmerili njihovo natančnost klasifikacije s statistično metodo cross-validation oz. prečnim preverjanjem klasifikacijske točnosti. Obravnavana in analizirana orodja, ki podpirajo metode strojnega učenja, so bila Weka, See5 in GATree. Rezultati analize so pokazali, da je od le teh najbolj natančno orodje za izvajanje klasifikacije program Weka.
Ključne besede: Strojno učenje, podatkovno rudarjenje, klasifikacija, orodja, Cross-validation.
Objavljeno: 19.09.2013; Ogledov: 1509; Prenosov: 453
Celotno besedilo (2,58 MB)
Traffic modelling and performance evaluation in the Kotor cruise portAndrej Škraba
, Branislav Dragović
, Davorin Kofjač
, Maja Škurić
, 2013, izvirni znanstveni članek
Opis: Simulation modelling of traffic in the Kotor cruise port is used for performance evaluation and optimization of the operational policy. During the summer, the traffic intensity of the cruise shipsŽ arrivals is higher than in other periods of the year. The increased traffic intensity causes congestion at anchorages, which can lead to significant dissatisfaction of the cruise ship operators and passengers. The simulation model, based on the port callingfrequency statistics and port tariff charges, is developed. First, the simulation model is validated against the real data. Second, several simulations are performed, where the scenarios of an extended main berth and of increased traffic intensity are evaluated in order to minimize congestion and to maximize revenue. Simulation results indicate significant queue reductions and a higher revenue, thus justifying the intention in the main berth extension.
Ključne besede: traffic modelling, port revenue, operational policy, simulation, model validation, Kotor cruise port
Objavljeno: 30.12.2015; Ogledov: 418; Prenosov: 2
Celotno besedilo (0,00 KB)
Disability in patients with chronic non-specific low back painZalika Klemenc-Ketiš
, 2011, izvirni znanstveni članek
Opis: Aim: To validate the Slovene version of the Oswestry Disability Index.
Method: A case series study of 129 adult patients with non-specific chronic low back pain attending a 10-day physical therapy programme. Visual Analog Pain Scale (VAS), Euroqol questionnaire (EQq), and Oswestry Disability Index (ODI) were used at the beginning and at the end of the physical therapy programme.
Results: Cronbach's a of the Slovene ODI was 0.876 at the beginning of the physical therapy programme and 0.901 at the end of the programme. All items of the questionnaire loaded on the same factor. A significant positive correlation existed between the Slovene ODI and VAS, EQ-5D and EQ-VAS at the beginning of the physical therapy programme (r=0.630, p<0.001; r=0.657, p<0.001; r=-0.510, p<0.001, respectively), as well as at the end of the programme (r=0.491, p<0.001; r=0.725, p<0.001; r=-0.648, p<0.001, respectively). The difference between ODI scores obtained at the first and at the second interview was 2.7 ± 5.2 (p<0.001), and the difference between ODI disability scores was 5.4 ± 10.5 (p<0.001). The effect size of the Slovene ODI questionnaire was 0.30.
Conclusions: The Slovene ODI questionnaire is a reliable and valid instrument for assessing outcomes of physical therapy in patients with chronic non-specific low back pain.
Ključne besede: disability evaluation, low back pain, physical therapy, validation studies, questionnaire
Objavljeno: 05.04.2017; Ogledov: 293; Prenosov: 38
Celotno besedilo (152,66 KB)
XML Schema metrics for quality evaluationMaja Pušnik
, Marjan Heričko
, Zoran Budimac
, Boštjan Šumak
, 2014, izvirni znanstveni članek
Opis: In XML Schema development, the quality of XML Schemas is a crucial issue for further steps in the life cycle of an application, closely correlated with the structure of XML Schemas and different building blocks. Current research focuses on measuring complexity of XML Schemas and mainly do not consider other quality aspects. This paper proposes a novel quality measuring approach, based on existing software engineering metrics, additionally defining the quality aspects of XML Schemas using the following steps: (1) definition of six schema quality aspects, (2) adoption of 25 directly measurable XML Schema variables, (3) proposition of six composite metrics, applying 25 measured variables and (4) composite metrics validation. An experiment was conducted using 250 standard XML Schemas collected from available e-business information systems. The results illustrate the influence of XML Schemaʼs characteristics on its quality and evaluate the applicability of metrics in the measurement process, a useful tool for software developers while building or adopting XML Schemas.
Ključne besede: XML, metrics, quality variables, quality index, evaluation, validation
Objavljeno: 06.07.2017; Ogledov: 185; Prenosov: 12
Celotno besedilo (765,37 KB)
Validation of agent-based approach for simulating the conversion to organic farmingČrtomir Rozman
, Andrej Škraba
, Karmen Pažek
, Davorin Kofjač
, 2017, izvirni znanstveni članek
Opis: Background and Purpose: The purpose of this study is to describe the principles of the development of parallel system-dynamics and agent-based models of organic farming for the case of Slovenia. The advantage of agent-based modeling is demonstrated by including geospatial information as an agent attribute. The models were compared by the validation, confirming the appropriate level of similarity.
Design/Methodology/Approach: Both system-dynamics and agent-based modeling approaches were applied. Statistical methods were used in the validation.
Results: The results of the validation confirm the appropriateness of the proposed agent-based model. Introducing additional attributes into the agent-based model provides an important advantage over the system-dynamics model, which serves as the paradigmatic example.
Conclusion: A thorough validation and comparison of the results of the system-dynamics and agent-based models indicates the proper approach to combining the methodologies. This approach is promising, because it enables the modeling of the entire agricultural sector, taking each particular farm into account.
Ključne besede: agent-based models, organic farming, system dynamics, validation, multimethod simulation
Objavljeno: 01.09.2017; Ogledov: 179; Prenosov: 18
Celotno besedilo (798,59 KB)
E-learning materials for social science studentsMarjan Krašna
, Tomaž Bratina
, 2014, izvirni znanstveni članek
Opis: Today's student population is rightfully categorized as digital natives. From the beginning of their education, they used ICT. The technological gadgets, internet, and social networks are like a glove to them. Such generation of students require more than just textbooks. In the 2013 University of Maribor establish a task force for e-learning materials development. The goal was to identify the optimal technological, didactical and financial approach to the long process of e-learning materials development. Members of the task force have many workshops presenting different views, acquired experiences from their previous projects, technological constraints, prediction, etc. During these workshops it was decided it would be the best to prepare different types of e-learning materials and test them by the students. From their feedbacks we could set the guidelines for large scale production. At the Faculty of Arts, Faculty of Natural Sciences and Mathematics and Faculty of Education students needs to acquire also the digital competences to become successful teachers. For special didactics study programs different types of e-learning materials were produced with different technological approach. The concept was to upgrade the previous deliverables (PowerPoint slides) and narrate them. Narrated slides would enable students to refresh their lectures and should be used as blended e-learning materials. Narrations were prepared in different format: textual narration, voice narration and video narration. But later it was decided that these types are not enough and a mix of narratives (integrated multimedia learning materials) could be used on individual slides depending on the content of the slide. Students first receive the lecture in the classroom and in the same week they need to study associated e-learning materials and write their review of corresponding e-learning materials. In their review students were required to log the required time for studying the e-learning materials; benefits and drawbacks; potential improvement of e-learning materials; and open text of impression of using e-learning materials. Effectiveness and outcomes were tested with the electronic quizzes. Students were highly motivated with these new types of learning materials and provide us valuable feedback. Most efficient were text narrated learning materials, but the favourite was integrated multimedia learning materials. Video and audio narrations take them more time to study since they were constrained with the speed of speech of the recorded lecturer.
Ključne besede: design, distance learning, e-learning, learning materials, validation, verification
Objavljeno: 15.12.2017; Ogledov: 81; Prenosov: 9
Celotno besedilo (1,47 MB)
Estimating piping potential in earth dams and levees using generalized neural networksXinhua Xue
, Xingguo Yang
, Xin Chen
, 2014, izvirni znanstveni članek
Opis: Internal erosion and piping in embankments and their foundations is the main cause of failures and accidents to embankment dams. To estimate the risks of dam failure phenomenon, it is necessary to understand this phenomenon and to develop scientifically derived analytical models that are simpler, easier to implement, and more accurate than traditional methods for evaluation of piping potential. In this study, a generalized regression neural network (GRNN) technique has been applied for the assessment of piping potential, as well, due to its ability to fit complex nonlinear models. The performance of GRNN has been cross validated using the k-fold cross validation method technique. The GRNN model is found to have very good predictive ability and is expected to be very reliable for evaluation of piping potential.
Ključne besede: piping, generalized neural network, cross validation, BP neural network
Objavljeno: 14.06.2018; Ogledov: 9; Prenosov: 0
Celotno besedilo (463,20 KB)