We present a comprehensive within-host mathematical model to quantify long-term hepatitis B virus (HBV) dynamics under pharmacological immune suppression and subsequent immune recovery. The model integrates nonlinear intracellular cccDNA production, ALT release, and both cytolytic and noncytolytic CD8+ T-cell responses within an eight-dimensional ODE framework. Using longitudinal HBV DNA, HBsAg, ALT, cccDNA, and CD8+ T-cell activation data from immunosuppressed rhesus macaques, parameters were estimated via the SAEM algorithm, yielding simulations that accurately reproduced prolonged viremia and cccDNA accumulation during immune suppression, followed by rapid viral decline after drug withdrawal. Long-term simulations revealed two distinct dynamical outcomes, viral clearance or persistent infection, emerging from variations in immune activation and killing efficacy. These results demonstrate how nonlinear intracellular regulation and immune-driven feedback collectively determine infection fate and establish a modeling framework applicable to broader classes of host–virus systems involving immune suppression and recovery.
A central problem in cross‐sectional stock‐return prediction is how to balance two competing forecasts built from the same firm characteristics: a stable long‐horizon forecast that averages predictive signals over many years and a responsive short‐horizon forecast that adapts quickly to recent market conditions. Existing combination rules set this balance using purely statistical criteria or a linear projection on macroeconomic uncertainty, implicitly assuming that uncertainty shifts the optimal balance by the same amount in all market states. This study examines whether uncertainty instead carries nonlinear and state‐dependent information about the optimal balance, and whether exploiting that information improves out‐of‐sample prediction. Using monthly returns and 94 firm characteristics for U.S. common stocks from January 1990 to December 2025, together with five uncertainty indices covering policy, financial, real, macroeconomic, and market‐implied uncertainty, we measure nonlinear and directional dependence with mutual information and transfer entropy and develop a nearest‐neighbor forecasting rule, MI–kNN–LASSO, which sets the balance by comparing the current uncertainty environment with historically similar states. All five indices relate to the optimal balance more strongly than linear correlations suggest, with macroeconomic uncertainty providing the most reliable predictive signal. The proposed rule forecasts the combination state far more accurately than the linear benchmark and earns a long–short return of 63 basis points per month with a smaller maximum drawdown. Our method offers investors an interpretable and adaptive way to combine return forecasts.
Financial markets are strongly nonstationary, making short-horizon foreign exchange decision support sensitive to estimation windows, feature design, and transaction costs. This paper proposes a transparent multiwindow ensemble framework in which interpretable multiregression experts use rolling-statistic features computed over different lookback windows. Time-scale diversity is combined with a supervisory layer that converts recent cost-aware utility scores into expert weights and selects execution thresholds on a meta window. The utility criterion is based on net pips after transaction costs and includes penalties for drawdown and turnover, while the execution rule incorporates volatility gating and minimum holding constraints. The framework is evaluated on synchronized one-minute quotes for 16 major currency pairs using a nonoverlapping walk-forward protocol. The benchmark set includes lag-based machine-learning models, LSTM, Transformer, Echo State Network, Bayesian model averaging, dynamic model averaging, stacking, and alternative multiexpert aggregation rules. In the main fixed-horizon experiment with τ=15 minutes, the proposed utility-weighted ensemble achieves the highest mean Sharpe ratio of 1.085 and a mean net profit of 312.3 pips. The results support utility-calibrated time-scale diversification as an auditable approach to FX decision support under nonstationarity.
The fractional Belousov–Zhabotinsky system (BZS) is an important reaction–diffusion model used to describe oscillatory chemical reactions, pattern formation, and memory-dependent dynamical processes arising in chemistry and related applied sciences. The main objective of this study is to investigate the analytical properties of the fractional BZS and to construct accurate approximate solutions for the corresponding nonlinear system. To achieve this goal, the locally compact Hausdorff property together with the compact-open topology is employed to establish the existence and uniqueness of solutions within an appropriate Banach space framework. In addition, the Hyers–Ulam stability of the proposed fractional model is analyzed, demonstrating the robustness of the obtained solutions under small perturbations of the initial data. Furthermore, a hybrid analytical technique combining the double Laplace transform and the Adomian decomposition method, called the double Laplace Adomian decomposition method (DLADM), is developed to derive approximate analytical solutions. The obtained results show that the proposed approach produces rapidly convergent solution series and yields excellent agreement with the corresponding exact solutions. The theoretical findings confirm the well-posedness and stability of the fractional BZS, while the numerical results demonstrate the effectiveness and accuracy of the DLADM in handling the nonlinear fractional model. The proposed framework may be useful for the analysis of nonlinear fractional reaction–diffusion systems arising in chemical kinetics, biological processes, and other applied mathematical models.
Seven of nine planetary boundaries have been transgressed and anthropogenic mass now exceeds all living biomass, yet existing collapse models assume ecological regeneration—contradicting tipping-point evidence—while integrated assessment models rely on smooth damage functions that preclude irreversibility. We introduce the dynamic biocompetitive ecological model (DBEM), a three-variable nonautonomous dynamical system coupling human population, technosphere mass, and a novel quantity—effective structural capacity—that represents the fraction of Earth’s life-support systems still functional. The DBEM’s central innovation is irreversible capacity decay: Once structural capacity is lost through technosphere metabolism, it does not regenerate. Calibrated to published 2025 data and recast in a nondimensional form, the model yields a single overshoot ratio of 1.23, confirming that the technosphere exceeds sustainably supported capacity by 23 percent. Two collapse thresholds emerge with fundamentally different time scales: an overshoot threshold, already crossed, that drives accelerating biosphere degradation, and a reproductive failure threshold, orders of magnitude more distant, that would constrain human population directly. Between them lies an intermediate regime—absent from all existing collapse and integrated assessment models—in which structural capacity is eroding but population dynamics remain unconstrained. This regime constitutes a policy-relevant window whose duration depends on the rate of capacity degradation and whose stabilization requires simultaneously reducing the material scale of the technosphere and arresting further biosphere erosion.
The study investigates the structural evolution and driving forces of the global foreign direct investment (FDI)–embodied carbon-transfer network from 2000 to 2019, utilizing a combination of social network analysis and econometric modeling. Leveraging the OECD Analytical Activities of Multinational Enterprises (AMNE) database and carbon-emission data, the research constructs a network of carbon flows embedded in FDI, identifying key transmission hubs and trends over 2 decades. The findings indicate that the network remained structured around a small number of major source economies, with the China aggregate surpassing the United States as the largest source of FDI-embodied carbon outflows by 2019. The study combines motif analysis with a temporal exponential random graph model (TERGM) to assess the determinants of network evolution. Compared with static ERGM or QAP regression, the TERGM evaluates the endogenous network structure, nodal attributes, dyadic covariates, and temporal change across a sequence of annual networks within a common framework. The results show that regional trade agreements are positively associated with tie formation across subperiods, while a formal interaction test indicates a stronger post-2009 association for international investment agreements. The normalized motif profiles also reveal the relative prominence of a recurring “carbon-transfer club” configuration, in which regional integration intensifies carbon-transfer ties among groups of economies. These findings extend aggregate and bilateral FDI-carbon accounting by identifying temporal and structural interdependence in global investment–related carbon transfers and underscore the need for policies that address the environmental implications of FDI while promoting sustainable economic integration.
Environmental sustainability remains a key policy and governance challenge for G20 economies, which contribute significantly to global production, energy consumption, and environmental pressures. This study examines how structural economic transformation, green technological innovation, energy use, and urbanization influence environmental performance across G20 countries over the period 2000–2024. Environmental performance is measured using the environmental performance index (EPI). To address cross-sectional dependence, slope heterogeneity, and mixed integration orders, the analysis employs advanced panel estimators, including CS-ARDL, augmented mean group (AMG), and common correlated effects mean group (CCEMG). The results indicate that economic complexity and green technology innovation significantly improve environmental performance, highlighting the importance of diversified production structures and innovation-oriented policies. Renewable energy consumption consistently enhances environmental outcomes, underscoring the role of clean energy policies in sustainability governance. In contrast, economic growth exerts a negative long-run effect, confirming the persistence of growth–environment trade-offs in the absence of effective regulation. Political stability shows a modest but positive impact, suggesting that stable governance environments support the implementation and continuity of environmental policies, though stability alone is insufficient without targeted interventions. Urbanization produces mixed effects, reflecting differences in urban planning capacity and policy coordination across countries. Overall, the findings emphasize the need for integrated governance strategies that combine political stability, green innovation, renewable energy expansion, and sustainable urban development to improve environmental performance in G20 economies.
The present work examines the dynamical features of a discrete-time Rosenzweig–MacArthur predator–prey system in which the predation rate is modulated by prey density. Through a combination of theoretical derivations and computational experiments, we establish the occurrence of period-doubling (PD) and Neimark–Sacker (NS) bifurcations, both of which culminate in chaotic regimes. The Lyapunov exponent spectrum corroborates this picture: Positive values mark the onset of chaos and reveal a pronounced dependence of the trajectories on parameter perturbations. Numerical simulations further reveal chaotic attractors and invariant closed trajectories. The Ott–Grebogi–Yorke (OGY) method is employed to stabilise unstable orbits to control chaos. When extending the analysis to a coupled network, we show that chaos appears beyond a certain critical coupling threshold. These insights advance understanding of predator–prey dynamics and chaos control in ecological systems.
This work examines a two-dimensional discrete Leslie-type predator–prey system in which the prey dynamics are subject to the Allee effect. The primary objective is to determine equilibrium states and examine bifurcation phenomena in the neighborhood of the positive fixed point, highlighting their ecological significance. The analysis of the interior equilibrium confirms that the model undergoes several essential dynamical changes, including one-parameter bifurcations together with period-doubling and Neimark–Sacker bifurcation mechanisms. To clarify the structure of these transitions, the associated nondegeneracy requirements are established and the critical normal form coefficients are explicitly evaluated. The chaotic responses and bifurcation-generated oscillations are subsequently suppressed through the combined use of state–feedback control and the Ott–Grebogi–Yorke (OGY) strategy. Biologically, the results emphasize the central role of the Allee effect in regulating predator–prey behavior. More precisely, a moderate Allee influence tends to stabilize the interacting populations, promote their coexistence, and strengthen the long-term viability of the ecosystem. Moreover, the bifurcation properties of the discrete predator–prey system are also explored in a coupled network framework. The numerical results show that when the coupling intensity surpasses a critical threshold, the network dynamics become chaotic and display intricate, irregular oscillatory patterns. In addition, stochastic simulations based on the Euler–Maruyama scheme were performed to examine system behavior under environmental variability, allowing the model to reflect diverse ecological conditions.
Surface electromyography (sEMG) plays a crucial role in decoding neuromuscular activity, with applications in prosthetic control and human–computer interaction. While extensive research has focused on hand gesture classification, discriminating identical hand movements performed at varying wrist angles remains underexplored. This study evaluates five feature reduction techniques, namely, principal component analysis (PCA), linear discriminant analysis (LDA), chi-square test, mutual information (MI), and recursive feature elimination with cross-validation (RFECV) for classifying five wrist-based hand movements. Two independent sEMG datasets were analyzed. Experimentally collected Dataset 1 comprised 17 subjects performing 5 repetitions of hand movements at wrist angles (+45°, +90°, 0°, −45°, and −90°) using the Delsys Trigno electromyography system. Dataset 2 (Ninapro DB1, Exercise B, Movements 13–17) comprised 27 subjects performing 10 repetitions of 5 distinct wrist and hand movements. Gradient boosting, LightGBM, random forest, SVM-RBF, and a 1D-CNN classifier were evaluated under different validation strategies, including subject-wise hold-out, subject-wise cross-validation, and movement-level cross-validation. RFECV consistently delivered strong performance across both datasets. Mean accuracies ranged from 77.73% under movement-level cross-validation to 90.11% under subject-wise hold-out, with best subject-wise accuracies reaching 95.38% on Dataset 1 and 96.35% on Dataset 2. In contrast, the 1D-CNN on raw EMG windows yielded only 36.64% mean accuracy, confirming the importance of feature engineering. These findings demonstrate that RFECV is a reliable and generalizable feature selection technique for sEMG-based movement classification, particularly for challenging tasks involving angular variations of identical hand movements.
Addressing the conflict between “centralized control and decentralized coordination” in the allocation of cognitive resources within the innovation ecosystem of platform-based manufacturing enterprises, as well as the governance dilemma of balancing short-term efficiency with long-term innovation, this paper constructs a multiagent interaction model encompassing internal “leadership–team–employees” and external “government–platform–partner enterprises–consumers.” Through MATLAB numerical simulations, the paper compares resource allocation efficiency under centralized and decentralized decision-making and reveals the characteristics of a double-threshold bifurcation and the optimal governance range. The results indicate that (1) excessive internal centralization inhibits innovation vitality, whilst decentralized decision-making, although enhancing innovation output, tends to lead to increased coordination costs; (2) external centralized governance is more conducive to the integration of cognitive resources, whereas the decentralized model suffers from significant fragmentation; and (3) the system exhibits an internal bifurcation threshold γI∗ = −0.403 and an external bifurcation threshold γE∗ = −0.374, which together form a three-segment bifurcation structure; (4) the optimal governance model is “internal decentralization + external centralization,” which simultaneously maximizes both grassroots innovation dynamism and external integration efficiency. This paper expands the application boundaries of distributed cognition and differential games in platform governance, providing a theoretical basis and quantitative decision-making tools for platform-based manufacturing enterprises to dynamically optimize the allocation of cognitive resources, and for governments to refine innovation incentive policies.
Dengue continues to be a major global health concern, disproportionately affecting vulnerable populations such as pregnant women and individuals with comorbidities. In this study, we propose a deterministic compartmental model that stratifies the human population into nine classes, including those with comorbid conditions and pregnancy, to capture the heterogeneities in disease transmission, progression, and outcomes. The model incorporates behavior-related exposure modifiers and hospitalization dynamics to reflect real-world complexities. We perform a thorough mathematical analysis of the model, including positivity, boundedness, and equilibrium analysis. Using a series of numerical experiments, we explore how variations in transmission rates, recovery rates, exposure risks, and hospitalization influence epidemic outcomes across subpopulations. The simulations reveal that comorbid individuals and pregnant women significantly alter the course and severity of dengue outbreaks, both directly and indirectly. Our findings underscore the importance of targeted interventions and subgroup-specific prevention strategies, offering novel insights to inform public health policy and dengue control programs.
In today’s digital era, cyber epidemics are evolving rapidly. They pose a growing global threat. A cyber epidemic involves malware that spreads across connected networks. In this paper, we propose a delayed variable-order fractional model to study the spread of malware. The model incorporates time-varying memory effects through Caputo’s variable-order fractional derivative. It also accounts for delayed antivirus response via a time-delay term. The system is solved numerically using a predictor–corrector scheme. To the best of our knowledge, this work presents the first delayed variable-order fractional cyber epidemic model that integrates both time-dependent memory and time-delay effects. The results indicate that as the fractional order decreases, the system dynamics become slower and more stable, with lower infection levels. In addition, the variable-order model provides a more flexible description of evolving memory effects in cyber epidemic systems.
Thunderstorms are frequent and destructive mesoscale phenomena in tropical and subtropical regions, posing major hazards through intense rainfall, lightning, strong winds, hail, and occasional tornadoes. In Bangladesh, premonsoon Kalbaishakhi (Nor’westers) are particularly damaging and remain difficult to forecast at local scales using computationally intensive numerical weather prediction approaches. This study aimed to develop and comparatively evaluate high-frequency data-driven forecasting frameworks for daily thunderstorm occurrence in Mymensingh, a thunderstorm-prone district in north-central Bangladesh and to assess their applicability for operational early warning. Daily thunderstorm frequency observations obtained from the Bangladesh Meteorological Department covering January 1, 1981, to December 31, 2024, were analyzed using a chronological training–testing split (80%–20%). Classical statistical models (ARIMA, ETS, TBATS, and GARCH), machine learning models (SVR, ANN, random forest, and Prophet), deep learning models (LSTMs), and sequential hybrid frameworks were evaluated using MAE, RMSE, MASE, and MAPE. Seasonal-Trend Decomposition using Loess (STL) was applied to address strong seasonality and intermittent zero-inflated behavior prior to hybrid modeling. Exploratory analyses revealed pronounced annual seasonality, minimal winter thunderstorm activity, and peak occurrences during April–June with substantial interannual variability. Among the nonseasonally adjusted models, ETS and ARIMA demonstrated comparatively reliable performance, while hybrid models generally improved forecasting accuracy. STL decomposition substantially enhanced predictive performance, with STL–LSTM achieving the lowest MAPE (2.25%) among individual models, whereas the STL–ARIMA–SVR hybrid model demonstrated the most balanced overall forecasting performance (MAE = 0.0274, RMSE = 0.4830, MASE = 0.2071, MAPE = 3.52%). For operational early warning assessment, STL–ARIMA–SVR achieved a probability of detection (POD) of 0.8435 and a critical success index (CSI) of 0.4157, indicating strong thunderstorm-event-detection capability under intermittent conditions. The findings demonstrate that seasonally decomposed hybrid forecasting frameworks can substantially improve localized thunderstorm prediction and support climate risk mitigation and operational early warning systems in Bangladesh. Future studies should incorporate multistation observations, atmospheric predictor variables, and real-time deployment evaluation to further improve forecasting robustness and operational applicability.
The transition toward carbon neutrality constitutes a complex adaptive process in which firms must continuously balance long-term environmental objectives against short-term financing constraints. While corporate green transformation is widely recognized as essential for sustainable development, the internal financial mechanisms shaping firms’ environmental strategies remain insufficiently understood. This study explores how controlling shareholders’ (CSs) equity pledge—a prevalent liquidity mechanism in emerging capital markets—generate nonlinear and threshold-dependent effects on corporate green transformation. We develop a constrained optimization model that embeds financing constraints and control-transfer risks into shareholders’ decision-making. The model formalizes a resource–risk trade-off and predicts a nonmonotonic relationship: Equity pledge initially facilitates green investment by easing liquidity constraints, but beyond a critical threshold, escalating control risk induces short-termism and crowds out substantive sustainability efforts. Using a balanced panel of Chinese A-share listed firms from 2014 to 2023, we empirically confirm a robust inverted U-shaped relationship between equity pledge intensity and green transformation. Further mechanism analyses reveal that financing constraints and digital transformation act as key transmission channels through which this nonlinear effect materializes. Heterogeneity tests indicate that the trade-off is most pronounced among non–state-owned firms, high-pollution industries, and firms with weak external governance. Extending the analysis, we show that although green transformation generally enhances firm value, excessive equity pledge undermines this benefit by amplifying governance risks. By integrating formal modeling with large-sample empirical evidence, this study demonstrates that corporate green transformation emerges from a complex interaction between liquidity incentive and risk constraint. The findings highlight equity pledge regulation as a critical leverage point for stabilizing the financial–environmental nexus in the transition toward low-carbon development.
Increasing international complexity and uncertainty challenges us to consider the future world state we create. The platform of higher education provides a space to equip the next generation of thinkers with perspectives, methods and approaches to navigate uncertain times through complexity theory. Despite the explicit focus on complexity within transdisciplinary approaches and the established link between embodied learning to enhanced student outcomes, research to date on how and why educators utilise embodied learning to enhance understanding of complexity within transdisciplinary higher education is still an emerging domain of scholarship. Therefore, the aim of this paper was to share how and why transdisciplinary educators from the UTS Transdisciplinary School approach the design, delivery and impact of embodied learning to support educating for complexity in a transdisciplinary co-joint degree. Six key insights are presented in this paper: (1) influences for implementation, (2) perceived student benefits, (3) perceived educator benefits, (4) implementation across a programme curriculum, (5) creating enabling conditions and (6) reflexive adaption across contexts. Through collaborative autoethnographic inquiry, this research surfaces embodied systems awareness as central to how we educate for complexity in transdisciplinary higher education. We demonstrate how embodied learning approaches support dynamic engagement with complex challenges, principles and characteristics and why this proves valuable in transdisciplinary higher education.
This paper proposes a novel network topology method that combines wavelet packet decomposition with a monotonic composite quantile regression neural network (MCQRNN). Using market data from 82 listed financial institutions spanning 2015–2025 and incorporating macroeconomic variables into the modeling framework, it aims to accurately identify the nonlinear features of multiscale systemic risk spillovers under different market conditions. The results reveal that the systemic risk spillover network of financial institutions exhibits clear multiscale frequency domain characteristics. In the risk spillover network, the banking sector consistently acts as a risk transmitter across all frequency bands, while other financial sectors serve as primary risk absorbers.
Eradicating poverty and hunger (SDG 1 and SDG 2) is central to sustainable development and environmental stability. This study proposes a nonlinear dynamical model describing the coupled evolution of poverty and hunger under awareness-driven mitigation strategies. The population is divided into socio-economically secure, poor, food-insecure, and co-burdened groups, together with a dynamic awareness variable that reduces transitions into adverse states. The model incorporates bidirectional coupling, endogenous reinforcement, partial recovery, and exogenous socio-economic shocks linked to climate variability and policy disruptions. Qualitative analysis establishes positivity, boundedness, and invariant regions. A poverty–hunger reproduction number derived via the next-generation matrix determines the local stability of the poverty–hunger-free equilibrium. A high-order numerical framework combining quasilinearization with shifted Legendre spectral collocation and domain decomposition is developed to simulate the model dynamics; this approach enables spectrally accurate, stable long-horizon simulations and provides a globally smooth solution approximation that supports the subsequent data-driven forecasting stage. Attention-enhanced reservoir computing is further employed to investigate the forecasting capability of data-driven models for reproducing the temporal dynamics generated by the proposed system. The framework supports evidence-based policies toward sustainable livelihoods and basic-needs security.
This research aims to optimally design a waste management network (WMN) for the agricultural sector, considering the circular economy (CE) and resilience dimensions. The principles of CE are integrated by repurposing agricultural waste for compost production, recovering it for biofuel production as renewable energy, reusing it as animal feed, and recycling it into paper. The concept of resilience is also incorporated through node criticality and the use of substitute products, ensuring the supply chain remains adaptable and robust in the face of disruptions. Taking into account sunflowers as a case product, this study proposes a multiobjective decision-making (MODM) model to design their WMN. On the other hand, the fuzzy robust optimization (FRO) method is implemented in the model to tackle the problem’s technical uncertainty. Finally, to address the proposed MODM model, a new solution approach called multichoice interactive weighted tchebycheff goal programming (MCIWTGP) is developed. A case study problem is then investigated along with a set of sensitivity analyses to validate the applicability of the suggested methodology. The results indicate that the proposed model is capable of achieving a high profitability level while maintaining a service level of 9.9 and limiting the number of critical nodes to only 7 within the designed WMN. Furthermore, increasing demand leads to a decrease in profit and service levels while increasing the number of critical nodes. In contrast, an increase in initial waste availability enhances profit and service levels and reduces the number of critical nodes. Moreover, expanding processing center capacity positively impacts profit and service levels and decreases network criticality.
