Search Results (252)

Ensemble methods have proven highly effective in enhancing predictive performance by combining multiple models. We introduce a novel ensemble approach, the Random Generalized Additive Logistic Forest (RGALF), which integrates generalized additive models (GAMs) within a random forest framework to improve binary classification tasks. Unlike traditional random forests, which rely on piecewise constant predictions in terminal nodes, RGALF fits GAM logistic regression (LR) models to the data in each terminal node, enabling it to capture complex nonlinear relationships and interactions among predictors. By aggregating these node-specific GAMs, RGALF addresses multicollinearity, enhances interpretability, and achieves superior bias–variance tradeoffs, particularly in nonlinear settings. Theoretical analysis confirms that RGALF achieves Stone’s optimal rates for additive models ($\mathcal{O}(n^{-2k/(2k+d)}$) under appropriate conditions, outperforming the slower convergence of traditional random forests ($\mathcal{O}\left(n^{-2/3}\right)$). Furthermore, empirical results demonstrate RGALF’s effectiveness across both simulated and real-world datasets. In simulations, RGALF demonstrates superior performance over random forests (RFs), reducing variance by up to 69% and bias by 19% in nonlinear settings, with significant MSE improvements (0.032 vs. RF’s 0.054 at $n=1000$), while achieving optimal convergence rates ($\mathcal{O}\left(n^{-0.48}\right)$ vs. RF’s $\mathcal{O}\left(n^{-0.29}\right)$). On real-world medical datasets, RGALF attains near-perfect accuracy and AUC: 100% accuracy/AUC for Heart Failure and Hepatitis C (HCV) prediction, 99% accuracy/100% AUC for Pima Diabetes, and 98.8% accuracy/100% AUC for Indian Liver Patient (ILPD), outperforming state-of-the-art methods. Notably, RGALF captures complex biomarker interactions (BMI–insulin in diabetes) missed by traditional models. Full article

Cognitive decline and dementia prevention are global priorities, with cognitive rehabilitation games showing potential to delay their onset or progression. However, these games require sufficient user engagement to be effective. Assessing the engagement through questionnaires is challenging for the individuals suffering from cognitive decline due to age or dementia. This study aims to explore the relationship between game difficulty levels, three EEG engagement indices (β/(θ + α), β/α, 1/α), and the self-reported flow state scale score during video gameplay, and to develop an accurate machine learning algorithm for the classification of user states into high- and low-engagement. Twenty-seven participants (nine older adults) played a stunt plane video game while their EEG signals were recorded using EPOCX. They also completed the flow state scale for occupational tasks questionnaire after the easy, optimal, and hard levels of gameplay. Self-reported engagement scores significantly varied across the difficulty levels (p = 0.027), with the optimal level yielding the highest scores. Combining the three EEG indices achieved the best performance, with F1 scores of 89% (within-subject) and 81% (cross-subject). Engagement classification F1 scores were 90% for young adults and 85% for older adults. The findings provide preliminary data that supports using EEG data for engagement analysis in adults and older adults. Full article

The strategic significance of software testing in ensuring the success of software development projects is paramount. Comprehensive testing, conducted early and consistently across the development lifecycle, is vital for mitigating defects, especially given the constraints on time, budget, and other resources often faced by development teams. Software defect prediction (SDP) serves as a proactive approach to identifying software components that are most likely to be defective. By predicting these high-risk modules, teams can prioritize thorough testing and inspection, thereby preventing defects from escalating to later stages where resolution becomes more resource intensive. SDP models must be continuously refined to improve predictive accuracy and performance. This involves integrating clean and preprocessed datasets, leveraging advanced machine learning (ML) methods, and optimizing key metrics. Statistical-based and traditional ML approaches have been widely explored for SDP. However, statistical-based models often struggle with scalability and robustness, while conventional ML models face challenges with imbalanced datasets, limiting their prediction efficacy. In this study, innovative decision forest (DF) models were developed to address these limitations. Specifically, this study evaluates the cost-sensitive forest (CS-Forest), forest penalizing attributes (FPA), and functional trees (FT) as DF models. These models were further enhanced using homogeneous ensemble techniques, such as bagging and boosting techniques. The experimental analysis on benchmark SDP datasets demonstrates that the proposed DF models effectively handle class imbalance, accurately distinguishing between defective and non-defective modules. Compared to baseline and state-of-the-art ML and deep learning (DL) methods, the suggested DF models exhibit superior prediction performance and offer scalable solutions for SDP. Consequently, the application of DF-based models is recommended for advancing defect prediction in software engineering and similar ML domains. Full article

The pervasive challenge of missing data in scientific research forces a critical trade-off: discarding incomplete observations, which risks significant information loss, while conventional imputation methods struggle to maintain accuracy in high-dimensional settings. Although approaches like multiple imputation (MI) and random forest (RF) proximity-based imputation offer improvements over naive deletion, they exhibit limitations in complex missing data scenarios or sparse high-dimensional settings. To address these gaps, we propose a novel integration of Multiple Imputation by Chained Equations (MICE) with Bayesian Random Forest (BRF), leveraging MICE’s iterative flexibility and BRF’s probabilistic robustness to enhance the imputation accuracy and downstream predictive performance. Our hybrid framework, BRF-MICE, uniquely combines the efficiency of MICE’s chained equations with BRF’s ability to quantify uncertainty through Bayesian tree ensembles, providing stable parameter estimates even under extreme missingness. We empirically validate this approach using synthetic datasets with controlled missingness mechanisms (MCAR, MAR, MNAR) and dimensionality, contrasting it against established methods, including RF and Bayesian Additive Regression Trees (BART). The results demonstrate that BRF-MICE achieves a superior performance in classification and regression tasks, with a 15–20% lower error under varying missingness conditions compared to RF and BART while maintaining computational scalability. The method’s iterative Bayesian updates effectively propagate imputation uncertainty, reducing overconfidence in high-dimensional predictions, a key weakness of frequentist alternatives. Full article

Colorectal cancer (CRC) is one of the leading causes of cancer-related deaths globally. Current treatment options including chemotherapy and targeted therapies face challenges such as resistance and toxicity. Cyclic guanosine monophosphate (cGMP)-specific phosphodiesterase 5 (PDE5) has emerged as a promising target for CRC therapy due to its role in regulating cellular processes like proliferation and apoptosis. This study focuses on the in silico design of a novel PDE5 inhibitor MS01 derived from the lead compound exisulind which has shown apoptotic effects but failed due to hepatotoxicity. Using Schrödinger’s Induced Fit Docking (IFD) and molecular dynamic simulations, MS01 was designed to enhance binding affinity and reduce toxicity. The docking studies showed that MS01 exhibits stronger interactions with key PDE5 residues, particularly Gln817 and Phe820. ADMET predictions indicate favorable pharmacokinetic profiles, with reduced risk of drug–drug interactions and improved bioavailability. Toxicity assessments revealed that MS01 and its analogs have moderate toxicity, with MS20 and MS21 demonstrating lower hepatotoxicity compared to exisulind. These findings suggest that MS01 has the potential to be a more effective and safer PDE5 inhibitor for CRC treatment pending further experimental validation. Full article

High-dimensional survival data, such as microarray datasets, present significant challenges in variable selection and model performance due to their complexity and dimensionality. Identifying important genes and understanding how these genes influence the survival of patients with cancer are of great interest and a major challenge to biomedical scientists, healthcare practitioners, and oncologists. Therefore, this study combined the strengths of two complementary feature selection methodologies: a filtering (correlation-based) approach and a wrapper method based on Iterative Bayesian Model Averaging (IBMA). This new approach, termed Correlation-Based IBMA, offers a highly efficient and effective means of selecting the most important and influential genes for predicting the survival of patients with cancer. The efficiency and consistency of the method were demonstrated using diffuse large B-cell lymphoma cancer data. The results revealed that the 15 most important genes out of 3835 gene features were consistently selected at a threshold p-value of 0.001, with genes with posterior probabilities below 1% being removed. The influence of these 15 genes on patient survival was assessed using the Cox Proportional Hazards (Cox-PH) Model. The results further revealed that eight genes were highly associated with patient survival at a 0.05 level of significance. Finally, these findings underscore the importance of integrating feature selection with robust modeling approaches to enhance accuracy and interpretability in high-dimensional survival data analysis. Full article

This paper proposes a novel two-stage ensemble framework combining Long Short-Term Memory (LSTM) and Bidirectional LSTM (BiLSTM) with randomized feature selection to enhance diabetes prediction accuracy and calibration. The method first trains multiple LSTM/BiLSTM base models on dynamically sampled feature subsets to promote diversity, followed by a meta-learner that integrates predictions into a final robust output. A systematic simulation study conducted reveals that feature selection proportion critically impacts generalization: mid-range values (0.5–0.8 for LSTM; 0.6–0.8 for BiLSTM) optimize performance, while values close to 1 induce overfitting. Furthermore, real-life data evaluation on three benchmark datasets—Pima Indian Diabetes, Diabetic Retinopathy Debrecen, and Early Stage Diabetes Risk Prediction—revealed that the framework achieves state-of-the-art results, surpassing conventional (random forest, support vector machine) and recent hybrid frameworks with an accuracy of up to 100%, AUC of 99.1–100%, and superior calibration (Brier score: 0.006–0.023). Notably, the BiLSTM variant consistently outperforms unidirectional LSTM in the proposed framework, particularly in sensitivity (98.4% vs. 97.0% on retinopathy data), highlighting its strength in capturing temporal dependencies. Full article

Forests are critical ecosystems that regulate climate, preserve biodiversity, and support human livelihoods by providing essential resources. However, they are increasingly vulnerable due to the growing impacts of deforestation and habitat fragmentation, which endanger their value and long-term sustainability. Assessing forest and deforestation fragmentation is vital for promoting sustainable logging, guiding ecosystem restoration, and biodiversity conservation. This study introduces an advanced approach that integrates the Local Connected Fractal Dimension (LCFD) with near real-time (NRT) land use and land cover (LULC) data from the Dynamic World dataset (2017–2024) to enhance deforestation monitoring and landscape analysis. By leveraging high-frequency, high-resolution satellite imagery and advanced imaging techniques, this method employs two fractal indices, namely the Fractal Fragmentation Index (FFI) and the Fractal Fragmentation and Disorder Index (FFDI), to analyze spatiotemporal changes in the forest landscape and enhance deforestation monitoring, providing a dynamic, quantitative method for assessing forest fragmentation and connectivity in real time. LCFD provides a refined assessment of spatial complexity, localized connectivity, and self-similarity in fragmented landscapes, improving the understanding of deforestation dynamics. Applied to Nigeria’s Okomu Forest, the analysis revealed significant landscape transformations, with peak fragmentation observed in 2018 and substantial recovery in 2019. FFI and FFDI metrics indicated heightened disturbances in 2018, with FFDI increasing by 75.2% in non-deforested areas and 61.1% in deforested areas before experiencing rapid declines in 2019 (82.6% and 87%, respectively), suggesting improved landscape connectivity. Despite minor fluctuations, cumulative deforestation trends showed a 160.5% rise in FFDI from 2017 to 2024, reflecting long-term stabilization. LCFD patterns highlighted persistent variability, with non-deforested areas recovering 12% connectivity by 2024 after a 38% reduction in 2019. These findings reveal the complex interplay between deforestation and landscape recovery, emphasizing the need for targeted conservation strategies to enhance ecological resilience and connectivity. Fractal indices offer significant potential to generate valuable insights across multiple spatial scales, thereby informing strategies for biodiversity preservation and adaptive landscape management. Full article

The extraction, processing and transport of crude oil in the Niger Delta region of Nigeria has long been associated with collateral environmental damage to the largest mangrove ecosystem in Africa. Oil pollution is impacting not only one of the planet’s most ecologically diverse regions but also the health, livelihoods, and social cohesion of the Delta region inhabitants. Quantifying and directly associating localised oil pollution events to specific petrochemical infrastructure is complicated by the difficulty of monitoring such vast and complex terrain, with documented concerns regarding the thoroughness and impartiality of reported oil pollution events. Earth Observation (EO) offers a means to deliver such a monitoring and assessment capability using Normalised Difference Vegetation Index (NDVI) measurements as a proxy for mangrove biomass health. However, the utility of EO can be impacted by persistent cloud cover in such regions. To overcome such challenges here, we present a workflow that leverages EO-derived high-resolution (10 m) synthetic aperture radar data from the Sentinel-1 satellite constellation combined with machine learning to conduct observations of the spatial land cover changes associated with oil pollution-induced mangrove mortality proximal to pipeline networks in a 9000 km² region of Rivers State located near Port Harcourt. Our analysis identified significant deforestation from 2016–2024, with an estimated mangrove mortality rate of 5644 hectares/year. Using our empirically derived Pipeline Impact Indicator (PII), we mapped the oil pipeline network to 1 km resolution, highlighting specific pipeline locations in need of immediate intervention and restoration, and identified several new pipeline sites showing evidence of significant oil spill damage that have yet to be formally reported. Our findings emphasise the critical need for the continuous and comprehensive monitoring of oil extractive regions using satellite remote sensing to support decision-making and policies to mitigate environmental and societal damage from pipeline oil spills, particularly in ecologically vulnerable regions such as the Niger Delta. Full article

High-dimensional data have attracted considerable interest from researchers, especially in the area of variable selection. However, when dealing with time-to-event data in survival analysis, where censoring is a key consideration, progress in addressing this complex problem has remained somewhat limited. Moreover, in microarray research, it is common to identify groupings of genes involved in the same biological pathways. These gene groupings frequently collaborate and operate as a unified entity. Therefore, this study is motivated to adopt the idea of a penalized semi-parametric Bayesian Cox (PSBC) model through elastic-net and group lasso penalty functions (PSBC-EN and PSBC-GL) to incorporate the grouping structure of the covariates (genes) and optimally perform variable selection. The proposed methods assign a beta process prior to the cumulative baseline hazard function (PSBC-EN-B and PSBC-GL-B), instead of the gamma process prior used in existing methods (PSBC-EN-G and PSBC-GL-G). Three real-life datasets and simulation scenarios were considered to compare and validate the efficiency of the modified methods with existing techniques, using Bayesian information criteria (BIC). The results of the simulated studies provided empirical evidence that the proposed methods performed better than the existing methods across a wide range of data scenarios. Similarly, the results of the real-life study showed that the proposed methods revealed a substantial improvement over the existing techniques in terms of feature selection and grouping behavior. Full article

G-quadruplex (G4) DNAzymes with peroxidase activities hold potential for applications in biosensing. While these nanozymes are easy to assemble, they are not as efficient as natural peroxidase enzymes. Several approaches are being used to better understand the structural basis of their reaction mechanisms, with a view to designing constructs with improved catalytic activities. Spermine alters the structures and enhances the activities of some G4 DNAzymes. The reported effect of spermine in shifting the conformation of some G4 DNAzymes from antiparallel to parallel has not been tested on multimeric G4 DNAzymes. In this study, we examined the effects of spermine on the catalytic activities of multivalent constructs of Bcl2, c-MYC, PS2.M, and PS5.M. Our findings show that spermine significantly improved the peroxidase activity of PS2.M, an antiparallel G4 DNAzyme, while there was no significant effect on c-MYC, which already exists in a parallel conformation. The addition of spermine led to a substantial increase in the initial velocity of PS2.M and its multimeric form, enhancing it by approximately twofold. Therefore, spermine enhancement offers promise in expanding the range of DNAzymes available for use as biosensing tools. Full article

High stocking density (HSD) and heat stress (HS) challenge broiler production. While antibiotics can mitigate the adverse effects of HS and HSD, their restricted use underscores the need to explore phytochemicals, particularly their combined effects under such conditions. This study investigated the influence of flavonoids, isoquinoline alkaloids, and their combinations as alternatives to bacitracin on growth performance, inflammatory status, gut morphology, and ceca microbiome in broilers raised under HSD and HS. A total of 2100 one-day-old male Ross 308 broiler chicks were distributed into 70 replicates, randomly assigned to one of seven dietary treatments and raised during the summer for 37 days. The treatments included normal stocking density (NSD, 10 birds/m²); HSD (15 birds/m²); HSD + 50 ppm of bacitracin (BCT); HSD + 300 ppm of flavonoids (FVNs); HSD + 80 ppm of isoquinoline alkaloids (IQAs); HSD + FVNs (1–10 days) and IQAs (11–37 days) (FVN-IQA); and HSD + IQAs (1–10 days) and FVNs (11–37 days) (IQA-FVN). The HS index reached or exceeded 160 during most of the experimental period. From 11 to 24 days of age, the HSD and BCT birds had lower body weight gain. The FVNs, IQAs, and their combinations decreased the corticosterone, IL-6, malondialdehyde, and heterophil–lymphocytes ratio compared to the HSD. Jejunal, ileal, and duodenal villi height/crypt depth ratio was lower in HSD than in other treatments except BCT. The α- and β-diversity, microbiota composition, and metabolic pathways were affected by treatment groups. Overall, FVNs, IQAs, and their combinations improved the growth performance, anti-inflammatory response, and gut health in broilers under HSD and HS, with the combinations exerting synergistic effects. Full article

Background: Benign prostatic hyperplasia (BPH) is a frequent condition in ageing men. Surgery is recommended for severe BPH symptoms and BPH-related complications. TURP is the reference standard for BPH surgery, but carries a risk of bleeding, which can lead to significant perioperative morbidity and mortality. To reduce bleeding during TURP, antifibrinolytic agents like tranexamic acid (TXA) have been studied. We aim to review the current evidence regarding TXA use during transurethral BPH surgery. Objective: This review aims to assess the efficacy and safety of tranexamic acid in reducing bleeding during transurethral benign prostatic hyperplasia surgery. Methods: Major clinical research databases such as PubMed, Cochrane Central Register of Controlled Trials, EBSCO, Scopus, Google Scholar, and Web of Science were searched from 2012 to 2022 for randomised controlled trials (RCTs) comparing the use of TXA to placebo in transurethral BPH surgery using the PICOS format. We included RCTs without language restrictions that assessed intraoperative blood loss, transfusion rates, haemoglobin levels, length of hospital stay, postoperative thromboembolic events, and 30-day perioperative mortality as outcomes. The quality assessment of the included studies was performed using the Cochrane risk-of-bias tool, RoB 2, for randomised studies. Results: A total of six RCTs, which included 456 patients, were eventually included in the meta-analysis. The results showed that tranexamic acid is beneficial in reducing blood loss and minimising changes in haemoglobin levels during transurethral resection of the prostate. However, it does not lessen the need for blood transfusions or shorten the hospital stay. Conclusions: Tranexamic acid is useful in decreasing blood loss and reducing changes in haemoglobin in patients undergoing transurethral resection of the prostate. Its utility during BPH surgery in low-resource settings where the latest haemostatic enucleation techniques, such as holmium and GreenLight laser enucleation, may not be readily available needs further evaluation. Full article

In this paper, we propose a novel method for producing image captions through the utilization of Generative Adversarial Networks (GANs) and Vision Transformers (ViTs) using our proposed Image Captioning Utilizing Transformer and GAN (ICTGAN) model. Here we use the efficient representation learning of the ViTs to improve the realistic image production of the GAN. Using textual features from the LSTM-based language model, our proposed model combines salient information extracted from images using ViTs. This merging of features is made possible using a self-attention mechanism, which enables the model to efficiently take in and process data from both textual and visual sources using the self-attention properties of the self-attention mechanism. We perform various tests on the MS COCO dataset as well as the Flickr30k dataset, which are popular benchmark datasets for image-captioning tasks, to verify the effectiveness of our proposed model. The outcomes represent that, on this dataset, our algorithm outperforms other approaches in terms of relevance, diversity, and caption quality. With this, our model is robust to changes in the content and style of the images, demonstrating its excellent generalization skills. We also explain the benefits of our method, which include better visual–textual alignment, better caption coherence, and better handling of complicated scenarios. All things considered, our work represents a significant step forward in the field of picture caption creation, offering a complete solution that leverages the complementary advantages of GANs and ViT-based self-attention models. This work pushes the limits of what is currently possible in image caption generation, creating a new standard in the industry. Full article