The paper analyzes the cardiovascular parameters of patients with heart disease. The aim of this study was to predict death in a patient with cardiovascular disease based on 12 parameters, using Random Forest and Logistic Regression algorithms. Parameters were tuned for both algorithms to determine the best settings. The most significant factors in the process predicted were found using the FEATURE SELECTION method of both algorithms. By comparative analysis of the obtained results, the highest accuracy of 90% was obtained using the Random Forest Algorithm.

The Canadian Fire Weather Index system [1] has been used worldwide by many countries as classic approach in fire prediction. It represents system that account for the effects of fuel moisture and weather conditions on fire behaviour. It numerical outputs are based on calculation of four meteorological elements: air temperature, relative humidity, wind speed and precipitation in last 24h. In this paper meteorological data in combination with Canadian Fire Weather Index system (CFWI) components is used as input to predict fire occurrence using logistic regression model. As logistic regression is a supervised machine learning method it’s based on user input in the form of dataset. Dataset is collected using NASA GES DISC Giovanni web-based application in the form of daily area-averaged time series in period of 31.7.2010 to 31.7.2020, it’s analysed and pre-processed before it is used as input for logit model. CFWI components values are not imported but calculated in run-time based on pre-processed meteorological data. As a result of this research windows application was developed to assist fire managers and all those involved in studying the fire behaviour.

The aim of this research is to implement Computer Vision technologies on existing published concept proposed by the same author in previous researches "Collaborative and Non-Collaborative Dynamic Path Prediction for Mobile Agents Collision Detection with Dynamic Obstacles". Author proposed usage of Computer Vision technologies in order to increase independency of single robotic units in the swarm. This new method and algorithm is based on analysis of behavior of human objects and its implementation in form of functional method and algorithm which can be used in mobile robotics. In prior research papers, several new terms are proposed and explained such as Metamorphous Hyperspace, Relevant predicted collision time, Coefficient of agility etc. The method implements human behavior in mobile robotics in a way it allows full decentralization of collision detection and ensures many other advantages starting from minimizing network traffic to simplifying inclusion of additional agents in relevant workspace. Algorithm requires a negligible amount of resources allowing mobile agents to exploit more resources for additional tasks. This method and algorithm can be implemented in all kinds of vehicles: ground, naval or airborne objects. Experimental model using Computer Vision technology OpenCV library is implemented and experimental result are described in this paper.

The fact that since March 2018 Agile is included in the Project Management Body of Knowledge (PMBOK), indicates the importance of this methodology in project management. Scrum as the one of the most accepted Agile forms emphasizes the importance of Scrum Master who has many responsibilities: teaching, coaching, mentoring the team, removing impediments, acting as a protector of the team, etc. However, many companies assign this role as additional work to a team member, a project manager or even to an executive. Guided by the impression that the role of Scrum Master is not fully recognized and appreciated, this paper will attempt to determine the potential risk of not having a dedicated Scrum Master within an agile team. As a research tool, we used interviews and online surveys with employees of companies that use Scrum.

In order to support modern business environment, software engineering has a constant tendency to increase productivity and thus the quality of the software. When recruiting, a biography can tell a lot about an individual, about expertise to do a certain job, but that is often not enough. For this reason, a model for prediction of work habits based on personality types has been introduced in this paper. Through the results of the research, by analyzing the employees of the FitSoft company (Bosnia and Herzegovina), a statement has been made of the existence of a certain correlation between the types of personality and the way they perform their tasks. The purpose of testing personality types is to determine how an individual performs their daily activities, and which features give them the advantage of choosing and performing certain jobs. A proposed model for predicting employees’ habits inside agile software teams based on personality estimates should present a clear insight of all factors that are related to employees, their productivity and business results.

Druga polovina prošlog vijeka okarakterisana je vrtoglavim razvojem računarskeindustrije koja je ubrzo postala osnovni preduslov razvoja svih sektora privrede.Uzimajući u obzir pomenuto, porasla je i potreba za stručnjacima koji će u segmentuposlovanja biti u stanju povezati tehničke i upravljačke zajednice. Rezultat je biopokretanje studija poslovne informatike, pod tim istim, sličnim ili različitim imenima,najčešće u okviru poslovnih škola kao studij prvog ciklusa. Međutim, razvojem upodručju poslovanja, a pogotovo kompjutinga, u novom mileniju se ovaj studij sve višepojavljuje kao samostalan sa tendencijom da postane ravnopravna organizacijskajedinica unutar hijerarhije univerziteta. Ovaj rad ima za cilj da predstavi teoretskuosnovu kurikuluma i praktičan primjer studija Poslovna informatika koji je rezultatzajedničkog rada Ekonomskog fakulteta i Fakulteta informacijskih tehnologijaUniverziteta „Džemal Bijedić“ u Mostaru.

The aim of this research is to finalize implementation of new method and algorithm of Collaborative and Non-Collaborative Dynamic Path Prediction for Mobile objects Collision Detection with Dynamic Obstacles in 2D and 3D Space. The method is based human behavior in collision detection with vehicles in real-life natural environment. Advantages of proposed method are full decentralization of the system, minimizing network traffic and simplifying inclusion of additional agents in the system. The proposed method is inspired by nature and implemented in mobile robotics. The method decreases uncertainty and increases predictability in collision detection with dynamic obstacles. Method allows implementation of fully functional algorithm which is tested in experimental environment and shows excellent results both in collaborative mode using exchange of coordinates as well as non-collaborative mode using OpenCV library for computer imaging and mobile objects tracking. The proposed algorithm is named Sliding Holt algorithm. This research paper should be considered as a part of series of research papers published earlier.

Higher education faces many challenges today, primarily those concerning the actual needs of the labor market. There is a high percentage of people with the university degree but considerably smaller percentage of those who are competent for solving actual problems that business world faces. This fact has been recognized as one of the drawbacks of Bologna process, more precisely of its traditional methods of teaching and learning, and was the main motivator for developing a new model of learning in this paper. It is a Competency-Based Education (CBE) model that aims to develop more competent students prepared for the labor market. It basically relies on the shortcomings of the traditional teaching model and introduces new elements to improve the entire teaching process, keeping those who have had positive outcomes in practice. The resulting model of competency based education has been validated at the Faculty of Information Technology in Mostar, where certain methods of learning that this model proposes are applied, and students are recognized as one of the best when it comes to applying their knowledge in practice.

Scarlet an Artificial Teaching Assistant is a personal digital assistant that has been developed with main aim to assist students in their learning process by ensuring fast and efficiently search of documents and learning materials. Scarlet is able to give an adequate response to a specific question based on knowledge gathered by an unique algorithm which enables her to recognize context during file and web page content search. After finding the most appropriate answer Scarlet seeks for student feedback in order to improve future search. The metric proposed is based on the power law which occurs in natural language, that is the Zipfian distribution[1]. It is designed to work for any spoken language although it might work on some better than other depending on the nature of the language, the structure, grammar and semantics. The method uses this metric to derive context from data and then queries the data source looking for the best match. The whole implementation is rounded off by a learning module which gives the system a learning curve based on users (students) scoring how relevant the output is among other parameters.

Scarlet an Artificial Teaching Assistant is a personal digital assistant that has been developed with main aim to assist students in their learning process by ensuring fast and efficiently search of documents and learning materials. Scarlet is able to give an adequate response to a specific question based on knowledge gathered by an unique algorithm which enables her to recognize context during file and web page content search. After finding the most appropriate answer Scarlet seeks for student feedback in order to improve future search. The metric proposed is based on the power law which occurs in natural language, that is the Zipfian distribution[1]. It is designed to work for any spoken language although it might work on some better than other depending on the nature of the language, the structure, grammar and semantics. The method uses this metric to derive context from data and then queries the data source looking for the best match. The whole implementation is rounded off by a learning module which gives the system a learning curve based on users (students) scoring how relevant the output is among other parameters. All the main algorithms and newly proposed metrics like the “contextual similarity” are presented in the same paper.

Discovering and sharing knowledge is considered to be a challenge faced by every modern academic institution. Recent research results have emphasized the fact that student skills and competencies cannot be measured in a simple way. Therefore, in this paper we present a list of parameters for discovering knowledge, based on available materials and human resources. The main goal of our research is to define an implementable model for measuring student skills and competencies in a real educational environment. Our model eliminates emerged shortcomings from our previous experience in working with distance learning platform and promises good results.