Search Results (33)

Accurate estimation of global horizontal irradiance (GHI) is relevant for characterizing solar resources in regions with limited measurement infrastructure. This study compared six machine learning models Polynomial Ridge Regression, Decision Tree, Random Forest, XGBoost, Artificial Neural Network, and K–Nearest Neighbors for the contemporaneous estimation of GHI at a five-minute resolution in an arid coastal climate. After quality control and restriction to daytime periods, 45,024 observations were analyzed using five external chronological blocks with an expanding-window scheme, generating 31,517 out-of-sample estimates. XGBoost achieved the highest $R^2$$(0.657\pm 0.192)$ and the lowest RMSE (108.01 ± 13.67 W m⁻²), whereas Random Forest yielded lower MAE, WMAPE, and MASE values. DM–HAC sensitivity analysis favored XGBoost under squared-error loss for six of the seven evaluated bandwidths, whereas no significant difference between XGBoost and Random Forest was found under absolute-error loss. None of the models achieved the nominal conformal coverage level of 90%; XGBoost showed the highest empirical coverage and the narrowest prediction intervals (PICP = 0.791; PINAW = 0.312). Predictor-set reduction improved the performance of four of the six algorithms. Overall, XGBoost and Random Forest exhibited complementary performance profiles, indicating that model selection should jointly consider predictive accuracy, temporal stability, predictor sensitivity, and uncertainty. Full article

Housing insecurity is one of the most urgent social problems in the United States, with eviction filings reflecting an important dimension of housing instability. This paper analyzes 75 monthly Florida eviction-filing observations from January 2019 through March 2025 to examine whether filing trajectories changed during and after pandemic-era tenant protections. Guided by an institutional temporality framework, we used descriptive analyses, a policy-guided interrupted time-series (ITS) model with Newey–West/HAC inference, a Fourier-seasonal dynamic regression with autoregressive errors, and exploratory segmented regression. The filing numerator covers all 67 Florida counties. Counts were the primary outcome; a secondary author-calculated rate used a fixed denominator of 2,794,102 renter-occupied housing units from the 2019–2023 American Community Survey 5-Year Table B25003. Mean monthly filings were 10,766 before March 2020, 6551 during March 2020–June 2021, and 11,701 during July 2021–March 2025. The primary ITS estimated an immediate March 2020 decline of 6854 filings (95% CI: 4541–9166 fewer; p < 0.001); the preferred AR(2) dynamic model also supported a large March decline (5957 fewer; p = 0.011). The July 2021 immediate change and direct tests of a higher post-period trajectory were not statistically significant. Overall annual seasonality was supported, but the apparent January and October peaks were not significant after adjustment. Florida filings therefore fell sharply around the onset of pandemic protections and subsequently returned toward the historical range, without statistically credible evidence that the post-moratorium trajectory became higher than the pre-pandemic trajectory. Full article

We establish an asymptotic theory for the Jones inverse-weighted kernel density estimator when length-biased observations form a strictly stationary short-range dependent sequence. The statistical difficulty is intrinsically composite: reciprocal weighting is singular at the origin, the normalizing mean is estimated from the same dependent sample, kernel localization shrinks with the bandwidth, and the centered summands form a row-wise stationary triangular array whose envelope diverges at rate $h_n^{-1}$. Under a non-negative compactly supported Lipschitz kernel, an inverse-moment condition, geometric $\alpha$-mixing, local regularity of the target density, and uniform local bounds on lagged bivariate densities, we prove strong uniform consistency on compact subsets of $(0,\infty )$ and, separately, the uniform stochastic bound $O_{\mathbb{P}}\{h_n^2+{(\log n/\left(n{h}_n\right))}^{1/2}\}$. A covariance-localization argument shows that the scaled serial-covariance contribution is $O\left\{h_n\log (1/h_n)\right\}=o\left(1\right)$, so the first-order pointwise variance coincides with that of the corresponding independent length-biased estimator. Pointwise and finite-dimensional Gaussian limits are obtained by an explicit big-block/small-block argument with off-diagonal covariance control. The ratio normalization is treated directly: its variance contribution, its product with the localized fluctuation, and its cross-covariance with that fluctuation are all negligible at the $\sqrt{n{h}_n}$ scale. We further derive first-order AMSE and AMISE criteria, their oracle bandwidths, and feasible pointwise studentization under undersmoothing. The numerical study separates oracle from data-driven bandwidth selection, evaluates full-ratio HAC and moving-block corrections, examines a Frank-copula Markov robustness design, and benchmarks the Jones estimator against an alternative length-biased estimator. The simulations support the first-order theory while demonstrating that persistent short-range dependence can remain consequential for finite-sample uncertainty. Full article

Honduras allocates a substantial share of GDP to public education, yet it continues to record weak human capital outcomes relative to its Central American neighbors. This study tests whether education-spending volume, or instead institutional quality and income inequality, better explains literacy outcomes in Honduras. Using annual data for 2012–2023, the period for which Honduras reports complete education-spending data to the World Bank, we estimate four parsimonious OLS specifications with Newey–West HAC standard errors. Regressors include education spending (% of GDP), government effectiveness, the Gini index, and log GDP per capita. Across all four specifications, education spending is not a robust predictor of adult literacy: it is non-significant in three of the four models (p > 0.10) and reaches conventional significance only in Model 4 (β = 0.606, p < 0.05), a result that is itself specification-dependent given the sign instability of the education-spending coefficient across models. Government effectiveness is positive and significant wherever included (β = 6.64 to 7.61, p < 0.05), and the Gini index is negative and significant wherever included (β = −0.38 to −0.54, p < 0.01), together explaining up to 74% of the variance in adult literacy (Adjusted R² = 0.740) with only twelve annual observations. These findings indicate that institutional quality and income inequality, rather than public spending levels, are the binding constraints on literacy outcomes in Honduras during the period examined. Increasing education budgets without parallel governance reform and redistribution is unlikely, on this evidence, to translate into improved literacy. Full article

This study examines the determinants of conditional volatility in energy-sector equity returns and their implications for sustainable energy resilience, energy security, and investment stability. Using a multi-stage econometric framework, the analysis investigates how global financial, commodity, and macroeconomic shocks are transmitted to volatility dynamics across heterogeneous energy companies. The dataset includes domestic and international firms, enabling a comparative assessment of volatility behavior and risk-transmission mechanisms under different market and institutional conditions. The empirical framework combines ARMA models for return dynamics, EGARCH/GARCH specifications for conditional volatility estimation, and OLS regressions with HAC standard errors to identify key determinants of volatility, including market indices, commodity prices, exchange rates, and major geopolitical and economic events. The findings reveal strong volatility persistence across all assets and asymmetric responses to market shocks in most cases. Global market conditions, particularly lagged MSCI World returns, significantly affect volatility, whereas commodity effects related to oil, gas, and coal remain heterogeneous across firms. Event-based regressors show that systemic shocks, including the COVID-19 pandemic and the European energy crisis, increase volatility, although geopolitical effects depend on firm-specific exposure. The results contribute to the sustainability literature by linking energy-sector financial volatility with market resilience, energy security, and stable investment conditions for the energy transition. Full article

This paper investigates whether prediction market settlements create incentives for temporary price pressure in Bitcoin spot markets. Using high-frequency data from February 2025 to January 2026 and actual contract-level data from Polymarket and Kalshi to identify economically relevant contract strikes, we document basis divergence between settlement oracle exchanges (Coinbase) and non-constituent exchanges (Binance) during expiry windows. Employing a difference-in-differences framework with month fixed effects, we find that a one standard deviation increase in strike proximity is associated with a 6.7 basis point constituent exchange price deviation during settlement windows. The estimate is precise under the baseline minute-level HAC specification, while exact paired-month permutation inference based on 12 settlement events yields $p=0.0256$; equal-weight event aggregation produces a larger negative estimate, indicating event heterogeneity. Monthly directional patterns are suggestive, though stricter event-level and above-versus-below-strike tests provide mixed evidence on directional asymmetry. Taken together, these findings provide reduced-form evidence consistent with settlement-related incentives and may raise broader settlement-design considerations for decentralized financial systems. However, the analysis does not directly observe trader intent or the underlying mechanism. Full article

This study investigates Bitcoin price forecasting using integrated sentiment analysis and gradient boosting within digital financial ecosystems. Two XGBoost models were developed using sentiment scores derived from Bitcoin news (2021–2024) and technical indicators, including GARCH-estimated volatility, Bollinger Bands, MACD, and RSI. The analysis uses 1042 daily Bitcoin observations and 10,025 sentiment records. Two model configurations were evaluated: one using only technical indicators and another incorporating daily aggregated sentiment scores. Model performance was assessed using Diebold–Mariano tests with Newey–West HAC variance estimation and walk-forward validation across 40 rolling windows. Contrary to expectations, sentiment features provided no statistically significant improvement over the technical-only model (p = 0.4888). Both models achieved identical test performance (R² = −0.16%). Walk-forward validation revealed substantial temporal instability (Mean R² = −126.30%, Std = 233.05%), highlighting the challenges of forecasting daily Bitcoin returns. Nevertheless, both XGBoost models significantly outperformed the random walk benchmark (DM statistic = −8.58, p < 0.0001), indicating that technical indicators capture exploitable market structure despite limited predictive accuracy for practical trading. These findings support the efficient market hypothesis and have implications for digital financial ecosystems integrating multimodal information. Full article

In complex socio-technical systems, human–AI collaboration is becoming fundamental to the processes of knowledge creation, content generation, and innovation. The existing innovation models typically consider only a single actor, the sole AI system, or a content artifact, and therefore do not capture the dynamics between these heterogeneous actors. This study introduces a Multimodal Graph Neural Network (MM-GNN), for modeling and analyzing innovation dynamics within Human–AI–Content (HAC) networks. The proposed framework is based on HAC networks as dynamic tripartite graphs, where human nodes, AI agent nodes, and content nodes are interconnected by edges representing interactions that evolve over time. Multimodal information, including text, image, code, and structured interaction traces, is merged by attention-based fusion, and multimodal dependency and evolution of interactions are modeled by relation-aware graph message passing and GRU-based temporal propagation. The innovation potential is realized as an upper-bounded composite score based on normalized novelty, entropy change, diffusion contribution, and human-rated creativity if available. The model is assessed as a composition of node-level classification and a regression model for innovation-level classification and estimation of continuous innovation potential. Experiments on synthetic HAC datasets and selected real-world AIGC corpora demonstrate that MM-GNN performs better than the graph learning and index-based baselines, with an average F1 score of 0.87, temporal stability ρ = 0.89, and lower regression error. The ablation and visualization analyses demonstrate that the multimodal fusion and temporal propagation are beneficial for representation quality, diffusion modeling, and interpretation. The results offer a mathematical and computational approach to the study of innovation as an emergent phenomenon of dynamic human, AI, and content interactions and lay the groundwork for additional validation on a more expansive socio-technical scale. Full article

The objective of this research is to analyze the relationship between ESG factors and financial development, measured by Domestic Credit to the Private Sector by Banks (DCB). The empirical analysis employs a balanced panel of 82 countries for the years 2016 to 2022, obtained from the World Bank database. The proposed econometric model incorporates multiple ESG factors, including environmental (E), social (S), and governance (G). The list of econometric models under consideration includes fixed effects, random effects, WLS (weighted least squares), dynamic panel, and fixed effects with HAC estimation. Based on the conducted tests, the fixed effects estimation method has been chosen because the presence of serial correlation, heteroskedasticity, and cross-sectional dependence suggests that other methods will not provide an adequate model. As a result, fixed effects enable obtaining reliable estimates regarding the relationships between ESG factors and DCB. In addition, a KNN (K-Nearest Neighbors) regression was used to analyze potential nonlinear effects of the factors. The results show the strong positive relationship between ESG factors and financial development. More specifically, the presence of clean energy sources is associated with a positive DCB, and the depletion of natural resources is negatively associated with DCB. Moreover, social and governance factors are positively associated with financial development. Full article

Climate change poses increasing challenges to livestock production in tropical regions, where rising temperatures, rainfall variability, and feed cost fluctuations affect productivity and economic stability. However, few studies have jointly quantified the effects of climatic and economic variables on swine production in tropical production systems, particularly in Brazil. This study examined the effects of maximum temperature, precipitation, and feed price on swine production density in Brazil’s main producing states. The analysis included Paraná and Rio Grande do Sul as the principal empirical base, while Mato Grosso was retained because of its strategic relevance but contributed only limited observations and was therefore interpreted more cautiously. Using monthly observations and a multiple linear regression model with heteroskedasticity- and autocorrelation-consistent (HAC, Newey–West) standard errors, we found that higher maximum mean temperatures were associated with lower production density: a 1 °C increase corresponded to an estimated decline of 1.34 × 10⁶ kg/km². Precipitation showed a positive association, with each additional millimeter corresponding to an increase of approximately 1.82 × 10⁵ kg/km², whereas a 0.173 USD/kg increase in feed price was associated with a reduction of about 6.2 × 10⁶ kg/km². Although the model explained only a modest share of monthly variation (R² = 0.162), the results suggest that climatic exposure and feed-cost pressure are relevant components of swine production dynamics in Brazil and should be considered in future climate-risk and agricultural planning. Full article

This study examines the impact of mobile wallet adoption on the profitability of a banking institution in Peru during the period 2021–2025, in a context of rapid digital transformation in financial services. The research adopted a quantitative, non-experimental, longitudinal, and explanatory design based on a single-bank case study. Mobile wallet adoption was measured through a synthetic index (IAD) constructed from five indicators using principal component analysis, while profitability was assessed through return on assets (ROA), return on equity (ROE), and aggregate monetary profitability. The effect of the IAD on profitability was estimated using generalized estimation equations with HAC-type robust standard errors. The results show that mobile wallet adoption exerts a positive and statistically significant effect on all three profitability indicators, with the strongest effect on aggregate monetary profitability, followed by ROE and ROA. These findings contribute to the literature by providing longitudinal evidence from an underexplored emerging economy and by showing that the financial effects of digital adoption differ according to the profitability measure considered. Overall, the study highlights the relevance of mobile wallet adoption as a strategic digital factor in banking performance within emerging financial contexts. Full article

Chieng Hac in northern Vietnam is expanding maize cultivation, intensifying water competition and soil erosion. This study mapped regional water balance and erosion using remote sensing and GISs by coupling the Thornthwaite–Mather (TM) water balance model with the Revised Universal Soil Loss Equation (RUSLE) at 12.5 m resolution. Land cover was classified into maize, tree crops, paddy, forest, and other types using Random Forest. The TM model used 2021 precipitation and temperature measurements to estimate evapotranspiration, surplus, and deficit, while the RUSLE quantified soil loss. Two scenarios were evaluated: a baseline reflecting existing land use and an adjusted case applying strip cropping on 10–20° maize slopes and converting maize to tree crops on slopes > 20°. Tree crop conversion increased evapotranspiration and prolonged seasonal deficits relative to maize, increasing water deficit from 1013.6 to 1022.2 mm/year. In contrast, the interventions reduced mean soil loss from 15.52 to 11.51 t/ha/year, with the largest decline in the 5–25 t/ha/year class. Residual hotspots persisted on steep slopes and near drainage lines. The integrated framework highlights trade-offs between erosion control and seasonal water availability, supporting slope-based land-use planning in upland agricultural systems. These findings offer guidance for slope-based land-use planning by indicating that intervention priorities should vary depending on slope conditions and local water availability. Full article

This study investigates whether external terms of trade (TOT) and mining-sector GDP in Peru share a stable long-run relationship. Although mining has played a central role in the country’s growth trajectory, its performance remains highly exposed to international price cycles, raising questions about its structural sustainability under persistent external shocks. Using quarterly data for 2001–2024, the analysis applies Johansen cointegration techniques and estimates a bivariate Vector Error Correction Model (VECM) to evaluate long-run co-movement and short-run adjustment dynamics. The results identify a single cointegrating relationship in which mining GDP acts as the primary adjustment variable, gradually correcting deviations from long-run equilibrium, while short-run TOT shocks do not exert direct contemporaneous effects on mining growth. The estimated speed of adjustment is low, suggesting a prolonged convergence process consistent with the capital-intensive and rigid structure of the mining sector. Robustness exercises—including estimation with heteroskedasticity and autocorrelation consistent (HAC) standard errors and an extended specification incorporating gross fixed capital formation—confirm the stability of the long-run relationship. These findings indicate that the structural sustainability of mining output depends on the interaction between external price dynamics and the sector’s capacity to adjust to persistent international shocks. The study concludes that, in the Peruvian case, structural sustainability in the mining sector is not determined solely by global price trends, but is also conditioned by domestic productive and institutional factors that govern the speed of adjustment in the presence of sustained external volatility. Full article

The paper examines the energy transition using Poland as a case study. The model was estimated based on annual data for Poland for the period of 1990–2024 (n = 35). The estimation was carried out using the OLS method with HAC correction, and the statistical significance of parameters was assessed using statistical tests. Based on econometric analysis, the impact was examined throughout the entire research period, with additional analysis of the structural break dummy for 2015. It was verified whether this impact had changed since 2015 compared to the earlier period. The data were used to calculate indicators, arranged in three groups: (1) capacity availability indicators (for the availability of the overall power system and for the renewable energy sources (RES)); (2) indicator of emission intensity (the indicator was defined as the ratio of total greenhouse gases emission to real GDP); (3) indicator of the economy’s energy intensity (the indicator was defined as primary energy consumption per unit of GDP). Annual summaries of these indicators constituted the input data for econometric modelling. The aim of the empirical analysis was to deepen the identification of mechanisms shaping greenhouse gas emission intensity by incorporating into the model indicators of generation capacity availability and measures of the economy’s energy intensity. The data collection based on constructed greenhouse gas emission intensity and energy intensity indicators of the economy enables the analysis of the increase in emission intensity regardless of the scale of the economy, in the system of power availability for the entire energy system, as well as for renewable energy sources. This approach makes it possible to move away from the analysis of absolute volumes toward a structural perspective that better reflects the real production capabilities of the power system as well as the efficiency of energy use in the economy. The results indicate that economic energy intensity is the dominant determinant of greenhouse gas emission intensity in Poland during the research period. The econometric analysis estimates show a positive and statistically significant relationship between energy intensity and emissions intensity, whereas generation capacity availability indicators—both for the total power system and for renewable energy sources—do not exhibit statistically significant effects. However, it was found that this impact was not constant throughout the entire period (β is 0.455 for pre-2015 and 0.325 for post-2015). Sensitivity analysis based on point elasticities reveals that a 1% increase in energy intensity of GDP leads to an increase in greenhouse gas emission intensity (by approximately 1.18% pre-2015 and 0.85% post-2015), whereas analogous changes in total capacity availability and RES availability are associated with substantially smaller effects (0.10% and 0.20%, respectively). These findings suggest that improvements in economy-wide energy efficiency played a more decisive role in reducing emissions intensity than short-term variations in generation capacity availability. Full article

This study aimed to investigate the effect of high-amylose corn starch (HACS) addition on the physicochemical properties and in vitro digestibility of an ultrasound-treated waxy rice flour (UWRF)–HACS blend system. As the proportion of HACS increased, the amylose content in the blends significantly increased (p < 0.05), while their water solubility index (WSI) and swelling power (SP) significantly decreased (p < 0.05). Additionally, the average particle size of the blends increased, and the surface of starch granules became smoother. Compared to UWRF, the blends did not generate new functional groups, but increased the starch’s relative crystallinity and short-range ordered structure. Rheological results indicated that the HACS-UWRF blends were mainly elastic and exhibited a typical weak gel system. In vitro digestibility results showed that the addition of HACS significantly increased the resistant starch (RS) content in the rice cakes (p < 0.05), while substantially reducing the hydrolysis index (HI) and estimated glycemic index (eGI) (p < 0.05). This study revealed the processing characteristics and gelatinization behavior changes in the HACS-UWRF blends. It provides a theoretical basis for the development of specialized flour for slow-glycemic rice cakes. Full article