This study presents the development and optimization of a microbial-enriched compost powder applied as a bioactive coating on porous fired-clay pellets, designed as a functional growing-media carrier derived from agricultural residues. The clay pellets were fabricated from locally sourced clay incorporating 15 wt% corn cob as a pore-forming additive and fired at 900 degrees C to generate a lightweight porous ceramic matrix suitable for microbial attachment. Key physical and mechanical properties of the pellets, including firing shrinkage, water absorption, apparent porosity, bulk density, and compressive strength, were quantitatively evaluated to verify structural integrity and carrier performance. The compost powder was enriched with three functionally distinct microbial groups-phosphate-solubilizing, antagonistic, and nitrogen-fixing microorganisms-and their proportions were optimized using Response Surface Methodology based on a Box-Behnken Design. Butterhead lettuce was employed as a model plant to validate the functional response of the integrated carrier-coating system. The optimized formulation (0.7% phosphate solubilizers, 0.235% antagonists, and 2.7% nitrogen fixers, w/w) produced a shoot dry weight of 2.83 g, in close agreement with the model-predicted value of 2.84 g (R-2 = 0.9938). The results demonstrate that biomass-modified fired-clay pellets can provide a mechanically stable and porous carrier platform for microbial compost coatings, while statistical optimization supports effective tuning of biological inputs within a fixed engineered matrix. The integrated system shows potential as a waste-derived functional material relevant to sustainable growing media and environmental engineering applications.
Accurate early-warning forecasting of corporate financial distress remains a critical challenge due to nonlinear financial relationships, severe data imbalance, and the high operational costs of false alarms in risk-monitoring systems. This study proposes an explainable voting ensemble framework for early-warning forecasting of corporate financial distress using lagged accounting-based financial information. The proposed framework integrates heterogeneous base learners, including Decision Tree, Neural Network, and k-Nearest Neighbors models, and is evaluated using financial statement data from 752 publicly listed firms in Thailand, comprising sixteen financial ratios across six dimensions: liquidity, operating efficiency, debt management, profitability, earnings quality, and solvency. To ensure robustness under imbalanced and rare-event conditions, the study employs feature selection, data normalization, stratified cross-validation, resampling techniques, and repeated validation procedures. Empirical results demonstrate that the proposed Voting Ensemble delivers a precision-oriented and decision-relevant forecasting profile, outperforming classical classifiers and maintaining greater early-warning reliability when benchmarked against advanced tree-based ensemble models. Probability-based evaluation further confirms the robustness and calibration stability of the proposed framework under repeated cross-validation. By adopting a forward-looking, early-warning perspective and integrating ensemble learning with explainable machine learning principles, this study offers a transparent and scalable approach to financial distress forecasting. The findings offer practical implications for auditors, investors, and regulators seeking reliable early-warning tools for corporate risk assessment, particularly in emerging market environments characterized by data imbalance and heightened uncertainty.
Micronutrient limitations, particularly boron (B) and zinc (Zn), may constrain peanut productivity in coarse-textured soils. This study evaluated the effects of B and Zn fertilization on growth, yield components, and agronomic efficiency (AE) of peanut grown in sandy soil using a 4 & times; 4 factorial design with B rates (0, 0.006, 0.012, and 0.017 g pot(-1)) and Zn rates (0, 0.019, 0.038, and 0.058 g pot(-1)) under greenhouse conditions. Zinc significantly affected shoot dry weight, with the highest biomass observed in the control treatment (18.52 g pot(-1)), but had limited influence on reproductive and yield traits. In contrast, B significantly enhanced pod- and seed-related parameters, increasing total seed weight by 46-51%, with the highest mean yield (16.78-16.82 g pot(-1)) recorded at 0.006-0.017 g B pot(-1). Principal component analysis explained 81.4% of total variance and identified pod number and seed-filling traits as the primary contributors to yield variation. Agronomic efficiency, expressed on a per-pot basis, was negative for Zn at all rates, whereas B AE reached 513.3 g seed increase g(-1) B applied at 0.006 g B pot(-1) and remained positive at 0.017 g B pot(-1). Overall, the stronger AE and yield response to B fertilization suggest that B availability was likely a more important constraint than Zn under the studied sandy soil conditions.
Generative AI (GAI) is profoundly reshaping teaching models, learning methods, and knowledge production systems in higher education. However, current applications remain fragmented and tool-based, lacking a unified theoretical framework to explain and guide its systemic role in higher education. To address this, this paper proposes a general framework for generative AI in higher education (GAI-UFHE). This framework constructs a systemic structural model of generative AI in higher education from four dimensions: cognitive enhancement, instructional reconstruction, knowledge co-creation, and governance constraints, and further analyzes its mechanisms of action and application paths. Research shows that this framework can effectively integrate the multiple roles of generative AI in instructional design, learning support, evaluation systems, and academic research, providing a theoretical foundation and practical pathway for the digital transformation of higher education.
This study developed and optimized a microbial-enriched compost powder designed to enhance the growth performance of Red Cos lettuce using Response Surface Methodology (RSM). Three plant growth-promoting bacteria bacteria- bacteria-Azotobacter sp., Bacillus spp., and Azotobacter vinelandii-were incorporated into compost powder and applied to high-porosity fired clay granules as the growth medium. A Box-Behnken Design was employed to evaluate individual and interactive effects of microbial concentrations on plant dry weight. ANOVA confirmed that the quadratic regression model was highly significant (p < 0.0001), with strong predictive capacity (R-2 = 0.9973; Adjusted R-2 = 0.9925) and an insignificant lack-of-fit. All microbial factors significantly influenced biomass production, with A. vinelandii exhibiting the most significant effect, followed by Azotobacter sp. and Bacillus spp. Response surface and contour plots revealed clear synergistic interactions among the variables and identified a distinct optimal region. The optimal formulation-0.19% Azotobacter sp., 1.84% Bacillus spp., and 0.39% A. vinelandii-yielded the highest observed dry weight (2.14 +/- 0.04 g), consistent with model predictions. Confirmation experiments showed that unbalanced microbial levels led to inferior growth responses, underscoring the importance of optimized ratios. The findings establish that microbial consortia integrated into compost powder and applied via porous clay granules can substantially improve plant biomass while supporting environmentally sustainable production systems. This optimized formulation provides a promising basis for further development of microbially enhanced growth media for leafy vegetables.
