In this paper we present details of a virtual tour and game for VR headset that are designed to investigate an interactive and engaging approach of applying VR to student recruitment for an undergraduate course. The VR tour employs a floating menu to navigate through a set of 360° panoramic photographs of the teaching environment and uses hotspot interaction to display further information about the course. The VR game is a fast-paced shooting game. The course information is embedded on cubes that the player needs to focus on and destroy. The game experience is expected to generate an engaging way to promote the course. This work in progress outlines the concept and development of the prototype, and discusses the next stages of testing in order to evaluate the effectiveness of applying VR to undergraduate student recruitment.

The global pandemic of the virus COVID-19 dramatically has impacted Higher Education Institutions (HEIs) in Bosnia and Herzegovina and worldwide. HEIs were forced to switch overnight to online lectures and exams without almost any teachers' and students' preparation and education. After one year of online classes at the University of Mostar (SUM), whether that kind of teaching impacts creativity. In order to find the answer to the research question, the authors used a questionnaire they developed and applied in 2015 to investigate students' perceptions about creativity at the University of Mostar. The research presented in this paper is limited to SUM students who have had online classes since March 2020. The primary research goal is to investigate whether there are any significant changes in students' perceptions of creativity compared to research from 2015. Namely, the authors investigate whether the enhanced use of IT and online platforms (Google Meet, SUMARUM – the University of Mostar’s variant of Moodle) affected students' creativity. This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.

This paper aims to provide an analysis of the profitability of audit firms in the Republic of Serbia during the period 2016-2018. The analysis is based on the data collected from the financial statements from all audit firms registered in the Republic of Serbia. The profitability analysis includes two goals. This paper will primarily provide a descriptive statistical analysis of the profitability of audit firms measured by return on assets and net income per employee. The following part of the research will answer the question of which factors have a significant impact on the profitability of audit firms. Profitability as a dependent variable is defined as return on assets and net income per employee, while independent variables include market share, current ratio, leverage, size, affiliation to the international network, etc. To answer this question, a regression statistics analysis will be conducted. The research result will indicate which factor can improve the performances of audit firms. This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.

Abstract Pesticide poisonings, intentional as well as accidental, are common, especially in undeveloped and developing countries. The goal of this study was to analyze the clinical presentation of patients hospitalized due to acute organophosphate (OPP) or carbamate pesticide (CP) poisoning as well as to analyze the factors that potentially influenced the severity and outcome of the poisonings. A retrospective cross-sectional study was performed. The age and gender of each patient were recorded, the type of ingested pesticide, whether the poisoning was intentional or accidental, number of days of hospitalization, the severity of the poisoning, and the outcome of the treatment (i.e., whether the patient survived or not). Clinical aspects of poisonings were analyzed, as well as the therapeutic measures performed. 60 patients were hospitalized due to acute OPP or CP poisoning, out of 51 (85.00%) were cases of intentional self-poisoning. The majority of patients were poisoned by OPPs (76.67%), in one-third the causative agent was malathion, followed in frequency by chlorpyrifos and diazinon. Dimethoate poisonings were manifested with the most severe clinical picture. A 70% or lower activity of reference values of acetylcholinesterase and butyrylcholinesterase was found in 50% and 58% of patients, respectively. The most common symptom was miosis (58.33%), followed by nausea and vomiting. Pralidoxime reactivated acetylcholinesterase inhibited by chlorpyrifos or diazinon, but not with malathion or dimethoate. Impairment of consciousness and respiratory failure, as well as the degree of acetylcholinesterase and butyrylcholinesterase inhibition, were prognostic signs of the severity of poisoning. The lethal outcome was more often found in older patients (t = 2.41, p = 0.019). The type of ingested pesticide significantly affects the severity and outcome of poisoning as well as the effectiveness of antidotes.

The authors of this study define a relationship with a best friend as a form of interpersonal relationship between two people who are close, based on mutual attraction, respect and recognition, within which there is support and protection, intimacy, satisfaction, enjoyment in one another's company, and successful resolution of problems. The purpose of this study was to establish the dimensions of a friendship relationship and test the structure of a questionnaire examining the quality of a friendship relationship in the category of "best friend". The research was conducted via a survey method on a sample of 316 students of 5th, 6th and 7th grades of elementary school. The factor structure of the questionnaire was studied by a combination of exploratory and confirmatory factor analysis. The analysis of the questiononaire's structure resulted in a 4-factor model with 17 items that meet the recent criteria in validation of research instruments. The extracted factors indicate the importance of intimacy, use of leisure time, emphasizing the person's value and protection. According to all important indicators, the final model was shown to be methodologically reliable as a simple instrument with a potential of wide practical application in research on quality dimensions of a relationship between best friends. Key words: factor analysis; intimacy; leisure time; value; protection. --- Autori ovoga rada definiraju prijateljski odnos s najboljim prijateljem kao oblik interpersonalnoga odnosa između dvije bliske osobe zasnovan na obostranoj privlačnosti, poštovanju i uvažavanju unutar kojeg dolazi do potpore i zaštite, intimiteta, zadovoljstva, uživanja u društvu i uspješnoga rješavanja problema. Svrha rada bila je utvrditi dimenzije prijateljskoga odnosa i testirati svojstva strukture upitnika kojim se ispituje kvaliteta prijateljskoga odnosa u kategoriji najbolji prijatelj. Istraživanje je provedeno na uzorku od 316 učenika 5., 6. i 7. razreda osnovne škole metodom anketiranja. Faktorska struktura upitnika istraživala se kombinacijom eksplorativne i konfirmativne faktorske analize. Analiza strukture upitnika rezultirala je 4-faktorskim modelom od 17 čestica koji ispunjava recentne kriterije u validaciji istraživačkih instrumenata. Ekstrahirani faktori ukazuju na važnost intimiteta, provođenja slobodnoga vremena, isticanja vrijednosti osobe te zaštite. Finalni se model prema svim bitnim indikatorima pokazuje kao metodološki pouzdan i za korištenje jednostavan instrument koji bi mogao imati široke praktične primjene među istraživačima i praktičarima u istraživanju dimenzija kvalitete odnosa među najboljim prijateljima.Ključne riječi: faktorska analiza; intimitet; slobodno vrijeme; vrijednosti; zaštita

A common vision from science fiction is that robots will one day inhabit our physical spaces, sense the world as we do, assist our physical labours, and communicate with us through natural language. Here we study how to design artificial agents that can interact naturally with humans using the simplification of a virtual environment. We show that imitation learning of human-human interactions in a simulated world, in conjunction with self-supervised learning, is sufficient to produce a multimodal interactive agent, which we call MIA, that successfully interacts with non-adversarial humans 75% of the time. We further identify architectural and algorithmic techniques that improve performance, such as hierarchical action selection. Altogether, our results demonstrate that imitation of multi-modal, real-time human behaviour may provide a straightforward and surprisingly effective means of imbuing agents with a rich behavioural prior from which agents might then be fine-tuned for specific purposes, thus laying a foundation for training capable agents for interactive robots or digital assistants. A video of MIA's behaviour may be found at https://youtu.be/ZFgRhviF7mY

This paper presents the research results on digital maturity in higher education institutions in Bosnia and Herzegovina (BiH). Empirical research was conducted among employees of eight public higher education institutions in spring 2020. Digital maturity was examined through seven dimensions. The results show that higher education institutions in BiH started the digitalization process more than five years ago. They have been continuously working on the digitalization of all business processes and activities. According to employees, on a scale from 1 to 5, digitalization of their higher education institutions is somewhere in the middle (between 3 and 4). The institutions have room for improvement in all dimensions. This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.

ABSTRACT Executive functions (EF) play a key role in child’s development as they are necessary prerequisites for everyday functioning and later academic success. Much research has been directed at examining whether EF are unidimensional or multidimensional construct. In this study, we tested two theoretically driven models and one mathematically driven EF model based on the Behavior Rating Inventory of Executive Functions – Preschool Edition (BRIEF-P). The sample for this study consisted of 102 children with intellectual disability (77 boys, 25 girls), aged 40–71 months (mean age- 62.1 months, SD- 7.6 months). Early childhood special education teachers completed the BRIEF-P Teacher Version. Confirmatory Factor Analysis was used to assess the validity of different EF models. The original BRIEF-P, with a second-order, three-factor model and one-factor model were not a good fit to the data. Mathematically driven one-factor model, with the addition of correlated errors between the scales of working memory and shift, and working memory and plan/organize was a good model fit. The current study indicates that EF differentiation begins to emerge at preschool age in children with intellectual disabilities. Understanding EF structure in children with intellectual disability will help create better intervention programs for this population.

Holographic massive multiple-input multiple-output (MIMO), in which a spatially continuous surface is being used for signal transmission and reception, has emerged as a promising solution for improving the coverage and data rate of wireless communication systems. To realize these objectives, the acquisition of accurate channel state information in holographic massive MIMO systems is crucial. This paper proposes a channel estimation scheme based on a parametric physical channel model for line-of-sight (LoS) dominated communication in millimeter and terahertz wave bands. The proposed channel estimation scheme exploits the specific structure of the radiated beams generated by the continuous surface to estimate the channel parameters in a dominated LoS channel model. Since the number of unknown channel parameters is fixed regardless of the number of antennas, the training overhead of the proposed scheme does not scale with the number of antennas. The simulation results demonstrate that the proposed estimation scheme significantly outperforms other benchmark schemes in a poor scattering environment.

Estimating individualized treatment effects (ITEs) from observational data is crucial for decision-making. In order to obtain unbiased ITE estimates, a common assumption is that all confounders are observed. However, in practice, it is unlikely that we observe these confounders directly. Instead, we often observe noisy measurements of true confounders, which can serve as valid proxies. In this paper, we address the problem of estimating ITE in the longitudinal setting where we observe noisy proxies instead of true confounders. To this end, we develop the Deconfounding Temporal Autoencoder, a novel method that leverages observed noisy proxies to learn a hidden embedding that reflects the true hidden confounders. In particular, the DTA combines a long short-term memory autoencoder with a causal regularization penalty that renders the potential outcomes and treatment assignment conditionally independent given the learned hidden embedding. Once the hidden embedding is learned via DTA, state-of-the-art outcome models can be used to control for it and obtain unbiased estimates of ITE. Using synthetic and real-world medical data, we demonstrate the effectiveness of our DTA by improving over state-of-the-art benchmarks by a substantial margin.

Smart lighting systems are based on sensors and tunable lighting systems are increasingly more prevalent, but collecting, analysing, and using the large data generated by these sensors are challenging. Inspired by research on evolutionary algorithms, it can be hypothesized that an adaptive lighting system can operate in real-time by adjusting its output through a decision-making algorithm based on data mining techniques. Such an adaptive lighting system requires two-order input from users; initial and continual. Initial conditions provide training to the system through human factors research investigating the interaction between humans and their environment. Continual conditions are provided by data collected through sensors in real-time, and they continuously feed into the decision-making algorithm to adjust the output to meet occupants’ biological and psychological needs. Research indicates that artificial intelligence techniques, such as evolutionary algorithms, can emerge as good candidates for this framework.