Search Results (10)

In bank-dominated financial systems, the accumulation of non-performing loans is a recognised source of systemic vulnerability, as correlated corporate defaults can erode bank capital, impair liquidity, and propagate stress across interconnected portfolios. Firm-level default detection thus constitutes a microprudential foundation of macroprudential stability: the reliable early identification of risky borrowers reduces both individual credit losses and the aggregate exposures that drive system-level fragility. Yet the use of structured data-mining pipelines for this task remains underexplored in Central and Eastern Europe. This study applies the CRISP-DM methodology to predict corporate loan default using data on 302 Slovak corporate borrowers, combining financial ratios from publicly available financial statements with selected company and loan-related information from internal bank records. Seven individual classifiers were developed and compared: decision trees (CART, CHAID, C5.0), logistic regression, discriminant analysis, and neural networks (MLP, RBF), together with a stacked ensemble based on their outputs. Model performance was evaluated using sensitivity, overall classification accuracy, and area under the ROC curve (AUC), with sensitivity treated as the primary criterion because of the asymmetric costs of misclassification in credit risk assessment. The results confirm that historical firm-level information provides a reliable basis for default prediction, with tree-based models consistently outperforming statistical and neural network approaches. The stacked ensemble achieved the strongest overall performance, whereas C5.0 and CHAID showed that interpretable classifiers can also deliver competitive predictive accuracy. A champion–challenger deployment architecture is proposed, in which the ensemble serves as the performance-oriented champion and interpretable models act as challengers; this arrangement contributes to the operational resilience of the credit-risk assessment process and aligns with macroprudential expectations of model governance, auditability, and explainability. The study offers a replicable methodological framework for integrating data-driven decision support into credit evaluation in comparable banking settings. Full article

Distributed denial-of-service (DDoS) attacks, a type of Denial-of-Service (DoS) attack in which the targeted server, service or network is overloaded with malicious traffic originating from various different sources with the aim of making such targets inaccessible for legitimate users, continue to pose a pertinent threat to the availability and integrity of organisational digital assets. While many studies have shown that machine learning models can provide high predictive accuracy in detecting such attacks, they often fail to evaluate the practicality of deploying such models to production. This study aims to address this gap by evaluating a considerable amount of pipelines based on five popular supervised classifiers for detecting DDoS attacks using the CICDDoS2019 dataset. The study employs a comprehensive methodology that combines both manual feature removal with automated encoding, scaling and feature selection integrated within pipelines. A total of 210 pipelines formed of five classifiers, three features selectors, two hyperparameter tuners and seven train–test splits were initially evaluated. Pipeline performance was assessed using both conventional and computational performance metrics. To identify the champion pipeline, a two-step approach was employed: composite scoring for shortlisting and statistical testing using Friedman and post hoc Nemenyi tests. The champion pipeline was shown to be Decision Tree coupled with Recursive Feature Elimination (with 20 features selected) and Grid Search hyperparameter tuning with a 90-10 train–test split. It achieved the most optimal balance of predictive capabilities and computational overheads, achieving an MCC of $0.993\pm 0.024$, training time of $0.194\pm 0.001$ s, inference time of $0.000998\pm 0.00008$ s, CPU time of $0.194\pm 0.008$ s and average memory usage of 15,167 ± 322 kilobytes across training and inference. The findings highlight the importance of a holistic and more nuanced approach when selecting a champion pipeline that is not only effective but also feasible for deployment in resource-constrained environments. Full article

Human-database interaction is inevitable in intelligent system applications, and accurately converting user-entered natural language into database query language is a critical step. To improve the accuracy, generalization, and robustness of text-to-SQL, we propose Nabil (a model for natural language conversion query language based on brain-inspired computing technology and a large language model). This model first leverages the spatiotemporal encoding capabilities of spiking neural networks to capture semantic features of natural language, then fuses these features with those generated by a large language model. Finally, a champion model is designed to select the optimal query from multiple candidate SQLs. Experiments were conducted on three database engines, DuckDB, MySQL, and PostgreSQL, and the model’s effectiveness was verified on benchmark datasets such as BIRD. The results show that Nabil outperforms existing baseline methods in both execution accuracy and effective efficiency scores. Furthermore, our proposed normalization and syntax tree abstraction algorithms further enhance the champion model’s discriminative capabilities, providing new insights for text-to-SQL research. Full article

The common oak (Quercus robur L.), though ecologically important and long-lived, has declined in Northern Europe due to historical land use and conifer-dominated forestry. In Latvia, where its distribution is limited, oaks support a rich biodiversity through features like tree-related microhabitats (TreMs) and diverse epiphytic communities. This study compared TreM and epiphyte diversity between planted mature oaks and relict champion oak trees across 16 forest stands. Epiphyte species were recorded using fixed-area frames on tree trunks, and TreMs were categorized following a hierarchical typology. Champion trees hosted significantly more TreMs and a greater variety, including 10 unique TreMs. While overall epiphyte diversity indices did not differ significantly, champion trees supported more specialist and woodland key habitat indicator species. The findings underscore the ecological value of legacy trees, which provide complex habitats essential for specialist taxa and indicators of forest continuity. Conserving such trees is vital for maintaining forest biodiversity and supporting ecosystem resilience in managed landscapes. Full article

Monumental trees, defined by their exceptional size, form, and age, are critical components of both cultural heritage and ecological systems. However, their conservation faces increasing threats from habitat fragmentation, climate change, and inadequate public policies. This review synthesized global research on monumental trees by analyzing 204 peer-reviewed articles published between 1989 and 2024 that were sourced from Web of Science and Scopus. Our bibliometric analysis highlighted Olea europaea and Castanea sativa as the most frequently studied species and identified a surge in publications after 2019, particularly from the USA, Italy, and Spain. Key research themes included conservation, biodiversity, and ecosystem services. The methodological approaches varied globally, encompassing ranking systems; GIS mapping; remote sensing; and non-invasive diagnostic tools, such as acoustic tomography and chlorophyll fluorescence. Conservation strategies discussed included vegetative propagation, cryopreservation, and legal risk management. Despite advances in these techniques, significant gaps remain in effectively addressing environmental pressures and integrating multidisciplinary approaches. We concluded that targeted, interdisciplinary strategies are essential to safeguard monumental trees as vital ecological and cultural landmarks. Full article

Agroforestry is recognized as a significant element in climate-smart agriculture due to its high potential for addressing food insecurity, climate change challenges, and ecosystem management. However, despite the potential benefits of agroforestry practices, the adoption by rural households in Sub-Saharan Africa is low. Adopting agroforestry practices requires understanding rural households’ socio-economic and socio-psychological factors. Hence, this study empirically examined the role of knowledge, attitudes, and perceptions in the uptake of agroforestry practices among rural households to better understand the adoption process. A sample of 305 households was obtained from three communities, namely, Swayimane, Umbumbulu, and Richmond, in KwaZulu-Natal province. Principal component analysis and a binary logistic regression model were employed to analyze the data. Knowledge, attitudes, and perceptions towards agroforestry were found to positively influence the adoption of agroforestry practices. The results also revealed that age, farming experience, education level, and land size were determinants of agroforestry adoption. Therefore, the study recommends that policymakers, extension officers, and climate change champions consider rural households’ socio-economic characteristics, knowledge, attitudes, and perceptions when designing agroforestry projects. Implementing training programs with practical demonstration is also recommended to increase awareness of the benefits of agroforestry practices and encourage rural households to protect on-farm trees and shrubs. Full article

Coaches’ player substitution strategies can change the tactical behavior and the final result of matches. This empirical study aims to describe the relationship/association of player substitution variables with the results of men’s UEFA Champions League matches during the 2018–2019 season. A total of 125 matches were analyzed using an ad hoc observation sheet created for this purpose. To measure the degree and strength of association between the variables studied, Chi-square and Cramer’s V tests were used, respectively. In turn, the Adjusted Standardized Residuals from the contingency tables were calculated to detect patterns of association. Likewise, a decision tree, in particular, the CHAID method, was used to predict and identify interactions. Player substitutions affect the final result and the findings prove it. An own team’s goal after 5–10 minutes of player substitution was win-related (positive impact) (90.40%, ASRs = 10.40), while an opposing team’s goal after 5–10 minutes of player substitution was loss-related (negative impact) (90.30%, ASRs = 10.30). Regardless of the match status, the positive impact increased the winning percentage. Furthermore, the match status was postulated as an indicator of the need to make player substitutions. It showed that player substitutions could determine the final result when teams were tying. On the other hand, the match location was not a differentiating factor between winning and losing teams, although the winning percentage was somewhat higher for home teams. Coaches could use this information to establish player substitution strategies that would allow them to perform at their best. Full article

Seedlings derived from two Ohio buckeyes (Aesculus glabra Willd.) trees, the National Champion growing in Illinois, USA, and the Ohio State Champion were sampled. The National Champion grows at the northern limits of the species’ native range. The Ohio State Champion grows in Huron County, Ohio, well within the eastern range of the species. We also sampled 40 adult trees growing in Ohio and Illinois. All trees were genotyped at six microsatellite loci. We found that 42 of the 44 sampled seedlings (95%) from the National Champion tree, collected over two seasons, exhibited only maternal alleles at all six microsatellite loci, indicating they were produced by self-fertilization. In contrast, all seedlings from the Ohio state champion tree (N = 48) exhibited non-maternal alleles, indicating they were produced by outcrossing. Our results suggest that when outcross pollen is not available, A. glabra will self-fertilize, but does so rarely or never when outcross pollen is available. Seed germination and early survival were similar for progeny of both champions, but seedlings from the National Champion show lower growth rates and higher mortality during a spring frost, possibly due to inbreeding depression. There was little evidence for genetic structure between trees sampled in Ohio and Illinois. Full article

Ubiquitous online learning is continuing to expand, and the factors affecting success and educational sustainability need to be quantified. Procrastination is one of the compelling characteristics that students observe as a failure to achieve the weaker outcomes. Past studies have mainly assessed the behaviors of procrastination by describing explanatory work. Throughout this research, we concentrate on predictive measures to identify and forecast procrastinator students by using ensemble machine learning models (i.e., Logistic Regression, Decision Tree, Gradient Boosting, and Forest). Our results indicate that the Gradient Boosting autotuned is a predictive champion model of high precision compared to the other default and hyper-parameterized tuned models in the pipeline. The accuracy we enumerated for the VALIDATION partition dataset is 91.77 percent, based on the Kolmogorov–Smirnov statistics. Additionally, our model allows teachers to monitor each procrastinator student who interacts with the web-based e-learning platform and take corrective action on the next day of the class. The earlier prediction of such procrastination behaviors would assist teachers in classifying students before completing the task, homework, or mastery of a skill, which is useful and a path to developing a sustainable atmosphere for education or education for sustainable development. Full article

Food and nutrition security remain at the top of development priorities in low income countries. This is especially the case for smallholder farmers who derive their livelihood from agriculture yet are often the most deprived. Inclusive agribusinesses have been championed as a key strategy to address local constraints that limit smallholders’ participation in regional and global value chains, thereby enhancing their livelihood, and food and nutrition security, accordingly. In this paper, based on a mixed method research approach, we explore the potential food and security contribution of inclusive agribusiness in Makueni county, Kenya. We focus on the smallholders’ constraints and needs, exploring the extent to which these are addressed by such purported pro-poor approach. First, using independent sample t-tests and a probit regression model, we explore who are able to participate in an ongoing intervention. We compare how participants and non-participants differ in terms of key socio-economic characteristics and establish which of these attributes are associated with successful integration into the business. Second, we again use independent sample t-tests to determine how the participants and non-participants compare in terms of their food and nutrition security. The household food and nutrition security is assessed with the conventional measurement tools: the household food insecurity access scale and the household food dietary diversity score. We find that participation in the inclusive agribusiness favors smallholder households with relatively higher production capacity in terms of better physical capital (land and number of mango trees, financial capital), access to loans, and human capital (age, education, and family size). Following income improvement, the participants’ household food security situation is significantly better than for non-participants. However, participation does not improve household dietary diversity, implying that improvement in income does not necessarily lead to better household nutrition security. To address the limitations of inclusive agribusiness, we propose policymakers and development actors to critically explore the contextual background prior to intervention design and implementation, and accordingly devise a broader approach for more inclusivity of the very poor and marginalized, and better food and nutrition security outcomes as a result. Given that not every smallholder could benefit from inclusive agribusiness for their food needs due to resource scarcity, alternative livelihood supports, including social protection programs and safety net plans, should be considered. Full article