We consider deterministic nonlinear discrete-time systems whose inputs are generated by policy iteration (PI) for undiscounted cost functions. We first assume that PI is recursively feasible, in the sense that the optimization problems solved at each iteration admit a solution. In this case, we provide novel conditions to establish recursive robust stability properties for a general attractor, meaning that the policies generated at each iteration ensure a robust <inline-formula><tex-math notation="LaTeX">$\mathcal {KL}$</tex-math></inline-formula>-stability property with respect to a general state measure. We then derive novel explicit bounds on the mismatch between the (suboptimal) value function returned by PI at each iteration and the optimal one. However, we show by a counterexample that PI may fail to be recursively feasible, disallowing the mentioned stability and near-optimality guarantees. We therefore also present a modification of PI so that recursive feasibility is guaranteed a priori under mild conditions. This modified algorithm, called PI<inline-formula><tex-math notation="LaTeX">$^{+}$</tex-math></inline-formula>, is shown to preserve the recursive robust stability when the attractor is compact. In addition, PI<inline-formula><tex-math notation="LaTeX">$^{+}$</tex-math></inline-formula> enjoys the same near-optimality properties as its PI counterpart under the same assumptions.

Simple Summary This study investigates lung cancer detection by combining metabolomics and advanced machine learning to identify small cell lung cancer (SCLC) with high accuracy. We analyzed 461 serum samples from publicly available data to create a stacking-based ensemble model that can distinguish between SCLC, non-small cell lung cancer (NSCLC), and healthy controls. The model has 85.03% accuracy in multi-class classification and 88.19% accuracy in binary classification (SCLC vs. NSCLC). This innovation relies on sophisticated feature selection techniques to identify significant metabolites, particularly positive ions. SHAP analysis identifies key predictors such as benzoic acid, DL-lactate, and L-arginine, shedding new light on cancer metabolism. This non-invasive approach presents a promising alternative to traditional diagnostic methods, with the potential to transform early lung cancer detection. By combining metabolomics and machine learning, the study paves the way for faster, more accurate, and patient-friendly cancer diagnostics, potentially improving treatment outcomes and survival rates.

The river channel cut into an alluvial substrate constantly adapts to changes in water flow and sediment transport. This paper presents a new approach to determine the magnitude of channel deformation based on the definition of dimensionless spatial parameters of channel deformation. The new approach’s applicability is evaluated by experimental studies on the Željeznica River in the Sarajevo Field in central Bosnia and Herzegovina. The morphological changes of the channel over a period of ten years were analyzed using dimensionless parameters of the channel geometry. A numerical analysis of changes in the channel of the Željeznica River was carried out using a mathematical model and a selected equation for sediment transport. Sensitivity analysis of the model parameters for channel deformation and statistical reliability analysis of the numerical model showed a good agreement between the modeled and observed values of channel deformation parameters during the analyzed period.

Numerical and pseudo data for pion electroproduction from four reaction channels, p(γ*, π0)p, p(γ*, π+)n, n(γ*, π−)p, and n(γ*, π0)n, from threshold up to W = 1.575 GeV are used to perform a single energy partial wave analysis. As a constraint, higher partial waves are taken from the MAID07 model and lower partial waves are fitted. It is demonstrated that truncated partial wave analysis in a full isospin can be obtained with this procedure. The results for photon virtuality Q2 = 0.5 GeV2 are presented. Electromagnetic Eℓ±, Mℓ±, and longitudinal Lℓ± multipoles are presented and discussed. In the first step, numerical data are generated, and the optimal number of lower partial waves required for a good data fit is determined. In the second step, the same procedure is applied using generated pseudo data.

Abstract This paper investigates the impact of artificial intelligence (hereinafter AI) on the accounting profession, emphasizing the need to adapt educational programs and business practices due to the emergence and development of AI technologies. A qualitative method with the help of a semi-structured interview was used to collect primary data. The participants were professionals in the accounting and information technology field who have the relevant knowledge and experience to consider this topic. The research aimed to discover the main problems that may arise when implementing artificial intelligence in the accounting profession, to determine how AI technologies affect the quality of financial reports, and whether education reform in accounting is needed due to the emergence of AI technologies. The results of the conducted research showed that AI technology will find its application in the accounting profession, that the quality of the financial report generated based on AI technology depends on the quality of the entered data, i.e. that the control function of accountants is of crucial importance, and that it is necessary to reform curricula in the context of exploiting the benefits of AI technologies.

In this article we investigate topography and electrical properties of potassium blue bronze thin films by the means of atomic force microscopy. Thin films were prepared by pulsed laser depositions on two types of substrates (Al2O3 and SrTiO3) at different substrate temperatures and partial oxygen pressures. On the basis of topography measurements, granular nature of the films was revealed. Surface roughness, as well as the size and the orientation of grains was determined. Films deposited on SrTiO3 substrates are quite homogeneous and the grains display ordering in preferential directions. Films on Al2O3 contain randomly oriented grains. Roughness of the films increases with the increase of the deposition temperature. Electrical measurements provided qualitative information regarding the electrical properties of the films.

Pesticides used in agriculture can contaminate foods like fruits and vegetables, posing health risks to consumers and highlighting the need for effective residue monitoring. This study explores aqueous two-phase systems (ATPSs) comprising phosphonium or ammonium ionic liquids (ILs) combined with ammonium sulfate as an alternative pre-treatment method for extracting and concentrating the pesticides clomazone, pyraclostrobin, and deltamethrin from strawberry samples. Liquid–liquid equilibrium measurements for each ATPS were conducted, followed by extraction experiments to determine the most efficient systems for pesticide extraction. Results showed that all three pesticides migrated effectively to the IL-rich phase across the tested ATPSs. For the most promising system, tetrabutylphosphonium salicylate ([TBP][Sal]) with ammonium sulfate, extraction efficiencies for each pesticide exceeded 98% under optimized conditions for parameters such as pH, temperature, and ATPS composition. Application of this ATPS to strawberries resulted in significant pesticide preconcentration, reaching mg/L levels suitable for detection by liquid chromatography. The method’s sustainability was supported by green chemistry metrics, with AGREEprep and AGREE scores of 0.68 and 0.55, respectively, underscoring its alignment with eco-friendly practices.

Abstract Introduction. Sports are frequently investigated in relation to alcohol consumption habits in adolescence, but few studies have reported various competitive- and recreational-sport factors as correlates of harmful alcohol consumption in youth. This cross-sectional study aimed to identify possible associations that may exist between sports factors and harmful drinking in adolescents, emphasizing the specific influence of participation in competitive and recreational sports. Material and Methods. The participants were 650 adolescents from Bosnia and Herzegovina (16-17 years of age, 313 females) who were tested for sociodemographic factors, competitive-sport factors, recreational-sport factors, physical activity levels (PALs), and outcome-alcohol drinking (obtained by Alcohol Use Disorders Identification Test – AUDIT). For the purpose of statistical analyses, the obtained AUDIT score was dichotomized (non-harmful vs. harmful drinking). Results. Harmful drinking was more common in males than in females (χ2 = 47.01, p < 0.001), and those adolescents who reported better competitive success in sports (χ2 = 41.81, p < 0.001). The PAL was greater in those males who did not report harmful drinking (t-test = 2.45, p < 0.05), with no significant difference in females or in the total sample. Logistic regressions were performed with sex as a confounding factor and revealed that competitive success in sports was significantly correlated with harmful drinking (OR = 1.32, 95% CI: 1.02-1.66, p < 0.05). Conclusions. This study highlighted the specific association between competitive sports success and harmful drinking in adolescents, indicating the need for a more detailed analysis of the background of the established association. Further analyses on the potential association between PAL and substance misuse in youth are warranted.

This research is aimed at the importance of the value of a healthy lifestyle, especially among the younger population (students), and as an indicator of the level of health, the optimal level of sugar in the blood and the normality of the Body Mass Index represent an established prerequisite for longer-term life and cell health. The researchers used the Mann- Whitney U test for differences between a total of 73 female students of the University of Tuzla (N-35 Faculty of Physical Education and Sport, N-38 Law). The research showed that the students of the Faculty of Physical Education and Sports who have a significantly higher number of physical activities in their daily activities have lower body mass index values and that showed a statistically significant difference between the students of these two faculties at the significance level of 99% (p<0.01 ) , also in the blood sugar level between these two groups of students, a statistically significant difference was shown at the level of 99% (p<0.01), and in this area of research, the female students of the Faculty of Physical Education and Sports have lower values (MEAN = 4.989 mmol/l) from students of the Faculty of Law (5.545 mmol/l). It is important to note that in both groups in both researched areas the results are in the normal criterion ranges, and based on the values of arithmetic averages, it can be seen that female students of the Faculty of Physical Education and Sports have lower values of blood glucose and body mass index, which would could lead to a connection with the specifics of the study programs, for example the scope of daily movement activities. Keywords: BMI, insulin, health

The aim of this research was to determine the physical activity of primary school students in the area of Herzegovina-Neretva Canton. The research was conducted on a sample of N=2589 elementary school students from I to IX grades. The obtained data show that out of the total number of respondents, there are more students who do not play sports (54.4%) than students who play sports (45.6%). Only among fourth grade there are more students who play sports (57.8%) than students who don't play sports (42.2%). In all other classes, there are more students who don't play sports. Keywords: physical activity, primary school, students, sports

Post-war Bosnia and Herzegovina remains a society in which ethno-nationalist rhetoric and war memories permeate public discourse. This paper investigates the use of metaphors in media texts written in English and the Bosnian, Croatian, and Serbian language on the topic of ‘two schools under one roof’. Writing about this peculiar construct within the country’s educational system tackles socio-political factors and their effects on the provision of primary and second-ary education. We describe how metaphors as cognitive devices are used to portray educational developments in this post-conflict country and the effects of their use on recipients. The paper points out cases of deliberate metaphor use, which influences recipients’ perspectives on the topic at hand. Keywords: conceptual metaphor; deliberate metaphor; segregation; MIPVU.

Abstract The paper focuses on analyzing key factors influencing the effectiveness of internal audit in the public sector of Bosnia and Herzegovina (BiH). Through multiple regression analysis (Ordinary Least Squares-OLS), the impact of five independent variables (competence of internal auditors, size of internal audit department, relationship between internal and external auditors, management support for internal audit, and independence of internal auditors), two control variables (organization’s sector and total number of employees in the organization), on the dependent variable (effectiveness of internal audit) was measured. Research results indicate that a significant portion (.2 = 45.00%, Adj .2 = 40.90%) of the variability in the effectiveness of internal audit can be explained by the variability of the five independent variables in the model. The largest positive impact on the effectiveness of internal audit is attributed to the size of the internal audit department and the independence of internal auditors. Conversely, management support for internal audit had the least impact, which contradicts our predictions and the results of previous research. The limited impact of management support on the effectiveness of internal audit in the public sector of BiH may stem from a lack of understanding regarding the role and importance of internal audit, as well as from perceiving it as merely a formal legal requirement without substantial value.

Abstract Current literature on the privacy paradox in personalised advertising lacks insight into how consumers’ knowledge of the data types used shapes their responses to these ads. Building on privacy calculus theory, theory of reasoned action, and signalling theory, this research explores how consumers’ knowledge of data types in personalised advertising influences their reactions. Multigroup path analysis examines differences in established relationships based on consumers’ data knowledge. The moderating effect of this knowledge in the relationship between perceived invasiveness and purchase intentions is also tested. Findings from a sample of millennials indicate that privacy concerns increase perceived invasiveness. However, multigroup and moderation analyses reveal that perceived invasiveness’ impact on purchase intentions varies with consumers’ data knowledge. Specifically, in the search history group, perceived intrusiveness negatively affects purchase intentions. These results underscore the situation-specific nature of the privacy calculus and assist advertisers in understanding consumer behaviour in response to personalised ads.