Search Results (2,575)

Extreme summarization generates very short summaries, typically a single sentence, answering the question “What is the document about?”. Although large language models perform well in text generation, fine-tuning them for summarization often requires substantial computational resources that are unavailable to many researchers. In this study, we present an effective method for instruction-tuning open decoder-only large language models under limited GPU resources. The proposed approach combines parameter-efficient fine-tuning techniques, such as Low-Rank Adaptation (LoRA), with quantization to reduce memory requirements, enabling training on a single consumer-grade GPU. We fine-tuned a pre-trained decoder-only model on the XSum dataset using an instruction-following format. Experimental results demonstrate that the proposed decoder-only approach achieves competitive performance on the XSum dataset under strict GPU memory constraints. On the full test set, the proposed 2G–1R pipeline attains ROUGE-1/2/L F1 scores of 46.0/22.0/37.0 and a BERTScore F1 of 0.917, outperforming the individual generator models in lexical overlap and semantic similarity. Evaluation was conducted using traditional overlap-based metrics (ROUGE) and semantic metrics, including BERTScore and G-Eval. While remaining competitive in ROUGE compared to strong encoder–decoder baselines, the pipeline consistently produces summaries with higher semantic quality. These findings demonstrate that large decoder-only language models can be efficiently fine-tuned for extreme summarization on limited consumer-grade hardware without sacrificing output quality. Full article

Free-space optical (FSO) communication is a promising alternative to radio-frequency (RF) and optical fiber systems due to its high data rates and large bandwidth. However, its performance is highly susceptible to atmospheric conditions such as fog, rain, snow, and haze. This paper analyzes fog-induced signal attenuation in the Morogoro and Mbeya regions of Tanzania using the Kim and Kruse attenuation models. To improve link performance, a quadrature amplitude modulation (QAM) multiple-input multiple-output (MIMO) FSO link was designed and analyzed using OptiSystem 22.0. In Mbeya, light fog conditions with 0.5 km visibility resulted in an attenuation of 32 dB/km, a bit error rate (BER) of 4.5 × 10⁻²³, and a quality factor of 9.79 over a 2.62 km link. In Morogoro, dense fog with 0.05 km visibility led to an attenuation of 339 dB/km, a BER of 1.12 × 10⁻¹⁵, and a maximum link range of 0.305 km. Experimental measurements were further conducted under clear, moderate, and dense fog conditions to systematically evaluate the FSO link performance. The results demonstrated that MIMO techniques significantly enhanced link performance by mitigating fog effects. Moreover, a dedicated application was developed to analyze transmission errors and evaluate system performance metrics. Additionally, a mathematical model of the FSO link was developed to describe and forecast the performance of the MIMO FSO system in atmospheric conditions impacted by fog. Full article

Water quality monitoring faces increasing challenges due to accelerating industrialization and urbanization, demanding accurate, real-time, and reliable prediction technologies. This study presents a novel ensemble learning framework integrating Gaussian Process Regression, Support Vector Regression, and Random Forest algorithms for high-precision water quality pH prediction. The research utilized a comprehensive spatiotemporal dataset, comprising 11 water quality parameters from 37 monitoring stations across Georgia, USA, spanning 705 days from January 2016 to January 2018. The ensemble model employed a dynamic weight allocation strategy based on cross-validation error performance, assigning optimal weights of 34.27% to Random Forest, 33.26% to Support Vector Regression, and 32.47% to Gaussian Process Regression. The integrated approach achieved superior predictive performance, with a mean absolute error of 0.0062 and coefficient of determination of 0.8533, outperforming individual base learners across multiple evaluation metrics. Statistical significance testing using Wilcoxon signed-rank tests with a Bonferroni correction confirmed that the ensemble significantly outperforms all individual models (p < 0.001). Comparison with state-of-the-art models (LightGBM, XGBoost, TabNet) demonstrated competitive or superior ensemble performance. Comprehensive ablation experiments revealed that Random Forest removal causes the largest performance degradation (+4.43% MAE increase). Feature importance analysis revealed the dissolved oxygen maximum and conductance mean as the most influential predictors, contributing 22.1% and 17.5%, respectively. Cross-validation results demonstrated robust model stability with a mean absolute error of 0.0053 ± 0.0002, while bootstrap confidence intervals confirmed narrow uncertainty bounds of 0.0060 to 0.0066. Spatiotemporal analysis identified station-specific performance variations ranging from 0.0036 to 0.0150 MAE. High-error stations (12, 29, 33) were analyzed to distinguish characteristics, including higher pH variability and potential upstream pollution influences. An integrated software platform was developed featuring intuitive interface, real-time prediction, and comprehensive visualization tools for environmental monitoring applications. Full article

The rapid development of Convolutional Neural Networks (CNNs) has markedly boosted the performance of object detection in remote sensing. Nevertheless, tiny objects typically account for an extremely small fraction of the total area in remote sensing images, rendering existing IoU-based or area-based evaluation metrics highly sensitive to minor pixel deviations. Meanwhile, classic detection models face inherent bottlenecks in efficiently mining discriminative features for tiny objects, leaving the task of tiny object detection in remote sensing images as an ongoing challenge in this field. To alleviate these issues, this paper proposes a tiny object detection method based on Normalized Gaussian Label Assignment and Multi-scale Hybrid Attention. Firstly, 2D Gaussian modeling is performed on the feature receptive field and the actual bounding box, using Normalized Bhattacharyya Distance for precise similarity measurement. Furthermore, a candidate sample quality ranking mechanism is constructed to select high-quality positive samples. Finally, a Multi-scale Hybrid Attention module is designed to enhance the discriminative feature extraction of tiny objects. The proposed method achieves 25.7% and 27.9% AP on the AI-TOD-v2 and VisDrone2019 datasets, respectively, significantly improving the detection capability of tiny objects in complex remote sensing scenarios. Full article

This study evaluates the performance of several machine learning models in predicting dissolved oxygen concentration in the surface layer of the Mar Menor coastal lagoon. In recent years, this ecosystem has suffered a continuous process of eutrophication and episodes of hypoxia, mainly due to continuous influx of nutrients from agricultural activities, causing severe water quality deterioration and mortality of local flora and fauna. In this context, monitoring the ecological status of the Mar Menor and its watershed is essential to understand the environmental dynamics that trigger these dystrophic crises. Using field data, this study evaluates the performance of eight predictive modelling approaches, encompassing regularised linear regression methods (Ridge, Lasso, and Elastic Net), instance-based learning (k-nearest neighbours, KNN), kernel-based regression (support vector regression with a radial basis function kernel, SVR-RBF), and tree-based ensemble techniques (Random Forest, Regularised Random Forest, and XGBoost), under multiple experimental settings involving spatial variability and varying time lags applied to physicochemical and meteorological predictors. The results showed that incorporating time lags of approximately two weeks in physicochemical variables markedly improves the models’ ability to generalise to new data. Tree-based regression models achieved the best overall performance, with eXtreme Gradient Boosting providing the highest evaluation metrics. Finally, analysing predictions by sampling point reveals spatial patterns, underscoring the influence of local conditions on prediction quality and the need to consider both spatial structure and temporal inertia when modelling complex coastal lagoon systems. Full article

Cooperative task allocation is one of the critical enablers for multi-Autonomous Underwater Vehicle (AUV) missions, but existing approaches often assume reliable communication that rarely holds in real underwater acoustic environments. We study here the performance and robustness of the Consensus-Based Bundle Algorithm (CBBA) for multi-AUV task allocation under realistically degraded underwater communication conditions with dynamically appearing tasks. An integrated simulation framework that incorporates a Dubins-based kinematic model with minimum turning radius constraints, a configurable underwater acoustic communication model (range, delay, packet loss, and bandwidth), and a full implementation of improved CBBA with new features, complemented by 3D trajectory and network-topology visualization. We define five communication regimes, from ideal fully connected networks to severe conditions with short range and high packet loss. Within these regimes, we assess CBBA based on task allocation quality (total bundle value and task completion rate), convergence behavior (iterations and convergence rate), and communication efficiency (message delivery rate, average delay, and network connectivity), with additional metrics on the number of conflicts during dynamic task reallocation. Our simulation results indicate that CBBA maintains performance close to the optimum when the conditions are good and moderate but degrades significantly when connectivity becomes intermittent. We then introduce a local-communication-based conflict resolution strategy in the face of frequent task conflicts under very poor conditions: neighborhood-limited information exchange, negotiation within task areas, and decentralized local decisions. The proposed conflict resolution strategy significantly reduces the occurrence of conflicts and improves task completion under stringent communication constraints. This provides practical design insights for deploying multi-AUV systems under weak underwater acoustic networks. Full article

This study presents a process-informed framework for assessing the differential credibility of diverse downscaling methodologies, including both statistical (simple and complex) and dynamical approaches. The methods evaluated include a convolutional neural network (CNN), the Locally Constructed Analog Method (LOCA), the Statistical DownScaling Model (SDSM), quantile delta mapping (QDM), simple interpolation with bias correction, and two regional climate models. As proof of concept, we apply the framework to evaluate the physical consistency of processes associated with wet-day occurrence at a site in the southern USA Great Plains. Additionally, we introduce a relative credibility metric that quantifies cross-method performance and outlines how this framework can be extended to other variables, regions, and downscaling applications. Results show that all downscaling methods perform credibly when the parent global climate model (GCM) performs credibly. However, complex statistical methods (CNN, LOCA, SDSM) tend to exacerbate GCM errors, while simpler methods (QDM, interpolation + bias correction) generally preserve GCM credibility. Dynamical downscaling, by contrast, can mitigate inherited biases and improve overall process-level credibility. These findings underscore the importance of process-based evaluation in downscaling assessments and reveal how downscaling model complexity interacts with GCM quality. Full article

Accurate classification of tomato ripening stages and quality estimation is pivotal for optimizing post-harvest management and ensuring market value. This study presents a rigorous comparative analysis of morphological and colorimetric features extracted via two state-of-the-art deep learning-based instance segmentation frameworks—Mask R-CNN and YOLOv8n-seg—and their efficacy in machine learning-driven ripening stage classification and quality prediction. Using 216 fresh-market tomato fruits across four defined ripening stages, we extracted 27 image-derived features per model, alongside 12 laboratory-measured physio-morphological traits. Multivariate analyses revealed that R-CNN features capture nuanced colorimetric and structural variations, while YOLOv8 emphasizes morphological characteristics. Machine learning classifiers trained with stratified 10-fold cross-validation achieved up to 95.3% F1-score when combining both feature sets, with R-CNN and YOLOv8 alone attaining 96.9% and 90.8% accuracy, respectively. These findings highlight a trade-off between the superior precision of R-CNN and the real-time scalability of YOLOv8. Our results demonstrate the potential of integrating complementary segmentation-derived features with laboratory metrics to enable robust, non-destructive phenotyping. This work advances the application of vision-based machine learning in precision agriculture, facilitating automated, scalable, and accurate monitoring of fruit maturity and quality. Full article

To address the scarcity of sub-meter remote sensing samples and structural inconsistencies such as edge blur and contour distortion in super-resolution reconstruction, this paper proposes SRCT, a super-resolution method tailored for sub-meter remote sensing imagery. The method consists of two parts: external structure guidance and internal structure optimization. External structure guidance is jointly realized by the structure encoder (SE) and structure guidance module (SGM): the SE extracts key structural features from high-resolution images, and the SGM injects these structural features into the super-resolution network layer by layer, achieving structural transfer from external priors to the reconstruction network. Internal structure optimization is handled by the backbone network SGCT, which introduces a dual-branch residual dense group (DBRDG): one branch uses window-based multi-head self-attention to model global geometric structures, and the other branch uses lightweight convolutions to model local texture features, enabling the network to adaptively balance structure and texture reconstruction internally. Experimental results show that SRCT clearly outperforms existing methods on structure-related metrics, with DISTS reduced by 8.7% and LPIPS reduced by 7.2%, and significantly improves reconstruction quality in structure-sensitive regions such as building contours and road continuity, providing a new technical route for sub-meter remote sensing image super-resolution reconstruction. Full article

This paper investigates the use of large language models (LLMs) as evaluators in multidimensional machine translation (MT) assessment, focusing on the English–Indonesian language pair. Building on established evaluation frameworks, we adopt an MQM-aligned rubric that assesses translation quality along morphosyntactic, semantic, and pragmatic dimensions. Three LLM-based translation systems (Qwen 3 (0.6B), LLaMA 3.2 (3B), and Gemma 3 (1B)) are evaluated using both expert human judgments and an LLM-based evaluator (GPT–5), allowing for a detailed comparison of alignment, bias, and consistency between human and AI-based assessments. In addition, a classroom calibration study is conducted to examine how rubric-guided evaluation supports alignment among novice evaluators. The results indicate that GPT–5 exhibits strong agreement with human evaluators in terms of relative quality ranking, while systematic differences in absolute scoring highlight calibration challenges. Overall, this study provides insights into the role of LLMs as reference-free evaluators for MT and illustrates how multidimensional rubrics can support both research-oriented evaluation and pedagogical applications in a mid-resource language setting. Full article

With the rapid development of Massive Open Online Courses (MOOCs), high dropout rates have become a major challenge, limiting the quality of online education and the effectiveness of targeted interventions. Although existing MOOC dropout prediction methods have incorporated deep learning and attention mechanisms to improve predictive performance to some extent, they still face limitations in modeling differences in course difficulty and learning engagement, capturing multi-scale temporal learning behaviors, and controlling model complexity. To address these issues, this paper proposes a MOOC dropout prediction model that integrates multi-scale convolution with a symmetric dual-attention mechanism, termed SDA-Net. In the feature modeling stage, the model constructs a time allocation ratio matrix (MRatio), a resource utilization ratio matrix (SRatio), and a relative group-level ranking matrix (Rank) to characterize learners’ behavioral differences in terms of time investment, resource usage structure, and relative performance, thereby mitigating the impact of course difficulty and individual effort disparities on prediction outcomes. Structurally, SDA-Net extracts learning behavior features at different temporal scales through multi-scale convolution and incorporates a symmetric dual-attention mechanism composed of spatial and channel attention to adaptively focus on information highly correlated with dropout risk, enhancing feature representation while maintaining a relatively lightweight architecture. Experimental results on the KDD Cup 2015 and XuetangX public datasets demonstrate that SDA-Net achieves more competitive performance than traditional machine learning methods, mainstream deep learning models, and attention-based approaches on major evaluation metrics; in particular, it attains an accuracy of 93.7% on the KDD Cup 2015 dataset and achieves an absolute improvement of 0.2 percentage points in Accuracy and 0.4 percentage points in F1-Score on the XuetangX dataset, confirming that the proposed model effectively balances predictive performance and model complexity. Full article

Automatically generated chains-of-thought (gCoTs) have become common as large language models adopt deliberative behaviors. Prior work emphasizes fidelity to internal processes, leaving explanatory properties underexplored. Our central hypothesis is that these traces, produced by highly capable reasoning models, are not arbitrary by-products of decoding but exhibit stable and practically valuable textual properties beyond answer fidelity. We apply a multidimensional text-evaluation framework that quantifies four axes—structural coherence, logical–factual consistency, linguistic clarity, and coverage/informativeness—that are standard dimensions for assessing textual quality, and use it to evaluate five reasoning models on the GSM8K arithmetic word-problem benchmark (~1.3 k–1.4 k items) with reproducible, normalized metrics. Logical verification shows near-ceiling self-consistency, measured by the Aggregate Consistency Score (ACS ≈ 0.95–1.00), and high final-answer entailment, measured by Final Answer Soundness (FAS0 ≈ 0.85–1.00); when sound, justifications are compact, with Justification Set Size (JSS ≈ 0.51–0.57) and moderate redundancy, measured by the Redundant Constraint Ratio (RCR ≈ 0.62–0.70). Results also show consistent coherence and clarity; from gCoT to answer implication is stricter than from question to gCoT support, indicating chains anchored to the prompt. We find no systematic trade-off between clarity and informativeness (within-model slopes ≈ 0). In addition to these automatic and logic-based metrics, we include an exploratory expert rating of a subset (four raters; 50 items × five models) to contextualize model differences; these human judgments are not intended to support dataset-wide generalization. Overall, gCoTs display explanatory value beyond fidelity, primarily supported by the automated and logic-based analyses, motivating hybrid evaluation (automatic + exploratory human) to map convergence/divergence zones for user-facing applications. Full article

The gut microbiome is a modifiable factor in cancer survivorship. Diet represents the most practical intervention for modulating the gut microbiome. However, diet–microbiome relationships in prostate-cancer survivors remain poorly characterized. We conducted a comprehensive analysis of diet–microbiome associations in 79 prostate-cancer survivors (ages 62–81) enrolled in a randomized exercise intervention trial, 59.5% of whom still have active metastatic disease. Dietary intake was assessed using the Diet History Questionnaire (201 variables) and analyzed using three validated dietary pattern scores: Mediterranean Diet Adherence Score (MEDAS), Healthy Eating Index-2015 (HEI-2015), and the Mediterranean-Dash Intervention for Neurodegenerative Delay (MIND) diet score. Gut microbiome composition was characterized via 16S rRNA sequencing. Dimensionality reduction strategies, including theory-driven diet scores and data-driven machine learning (Random Forest, and Least Absolute Shrinkage and Selection Operator (LASSO)), were used. Statistical analyses included beta regression for alpha diversity, Permutational Multivariate Analysis of Variance (PERMANOVA) for beta diversity (both Bray–Curtis and Sørensen metrics), and Microbiome Multivariable Associations with Linear Models (MaAsLin2) with negative binomial regression for taxa-level associations. All models tested interactions with exercise intervention, APOLIPOPROTEIN E (APOE) genotype, and testosterone levels. There was an interaction between MEDAS and exercise type on gut alpha diversity (Shannon: p = 0.0022), with stronger diet–diversity associations in strength training and Tai Chi groups than flexibility controls. All three diet-quality scores predicted beta diversity (HEI p = 0.002; MIND p = 0.025; MEDAS p = 0.034) but not Bray–Curtis (abundance-weighted) distance, suggesting diet shapes community membership rather than relative abundances. Taxa-level analysis revealed 129 genera with diet associations or diet × host factor interactions. Among 297 dietary variables tested for cognitive outcomes, only caffeine significantly predicted Montreal Cognitive Assessment (MoCA) scores after False Discovery Rate (FDR) correction (p = 0.0009, q = 0.014) through direct pathways beneficial to cognitive performance without notable gut microbiome modulation. In cancer survivors, dietary recommendations should be tailored to exercise habits, genetic background, and hormonal status. Full article

Background: The increasing integration of Environmental, Social, and Governance (ESG) factors into financial decision-making has prompted debate over their impact on corporate credit risk. While many studies suggest that ESG performance may enhance firms’ resilience, empirical evidence remains mixed due to data inconsistency and methodological heterogeneity and differences in time horizons over which ESG effects materialise. Methods: The study investigates the relationship between ESG performance and credit risk using a panel of European firms from 2020 to 2024, a phase highly characterised by substantial macroeconomic shocks. The Altman Z-score serves as a proxy for default risk, while ESG data are sourced from Refinitiv Eikon. Four fixed-effects panel regressions are estimated: a baseline model using aggregate ESG scores, an extended model with financial controls, and disaggregated and sector-specific models. Results: The findings indicate that ESG scores—either aggregated or by pillar—show limited statistical significance in explaining variations in the Z-score. In contrast, financial variables such as solvency, liquidity, and cash flow ratios display strong, positive, and significant effects on credit stability. Some heterogeneous sectoral effects emerge: social factors are positive in technology, while governance has a negative impact in basic materials. Conclusions: ESG initiatives may not yield immediate improvements in default risk metrics, particularly over short and crisis-dominated periods, but could enhance financial resilience over time. Combining ESG information with traditional financial ratios remains essential; the results underscore the importance of consistent and high-quality ESG disclosure to reduce measurement error and enhance comparability across firms. Full article

Citrus, as a globally important economic crop, requires accurate assessment of premium fruit production potential for precise orchard management and enhanced economic benefits. This study develops a method for assessing the production potential of premium citrus using UAV-based multispectral imagery and ground data. Taking citrus orchards in Wuming District, Guangxi, China, as the experimental area, this study investigates techniques for assessing the production potential of premium fruit at the canopy scale of citrus trees in southern hilly regions, aiming to rapidly predict the quality production potential of citrus before fruit ripening. The methodology involved the following: (1) Segmenting the study area using a Digital Surface Model (DSM) and extracting individual tree canopies by integrating NDVI with a marker-controlled watershed algorithm. Canopy fruit boundaries were identified using the NPCI index. (2) Selecting key assessment indicators—NDVI, TCAVI, REOSAVI, canopy area, and canopy fruit area—through correlation analysis with nutritional quality metrics. (3) Establishing threshold levels for these indicators and constructing a production potential assessment model. Experimental results demonstrated an individual tree identification accuracy (precision) of 98.75%, a recall of 98.47%, and an F-score of 98.61%. Canopy area extraction achieved a coefficient of determination (R²) of 0.869 and a root mean square error (RMSE) of 0.489 m². The overall accuracy for production potential assessment reached 85.11%. This study provides a new approach for using UAV multispectral technology to non-destructively assess the production potential of premium citrus in the hilly regions of southern China, offering technical support for precise orchard management. Full article