Abstract Geopolymers offer a low-carbon alternative to ordinary portland cement (OPC); yet their industrial adoption is hindered by complex multivariate design spaces exceeding 10 6 possible formulations and reliance on empirical optimization. This study integrates mechanochemical activation (MCA), central composite design (CCD), ensemble machine learning (ML), and explainable artificial intelligence (XAI) to systematically optimize ground granulated blast furnace slag (GGBS)-based geopolymer pastes and derive quantitative design guidelines. Fifteen CCD experiments combined with 1,036 literature records (total n = 1,051 ) were used to train and validate seven predictive models. Among seven models evaluated, categorical boosting (CatBoost) achieved the best ensemble performance ( R 2 = 0.939 , root-mean-square error ( RMSE ) = 2.27 MPa , MAPE = 0.37 % ) and was selected for posthoc explainability analysis; support vector machine (SVM) achieved higher test R 2 (0.971) but is incompatible with TreeSHAP-based interpretation. Multiscale explainability analysis via SHapley Additive exPlanations (SHAP), partial dependence plots (PDPs), and individual conditional expectation revealed that temporal parameters dominate strength evolution, with curing time accounting for 67.3% of total model explainability and temporal-chemical interactions collectively explaining 96.4% of predictive variance. Precise optimal ranges were identified: activator molarity 10–12 M, SS/SH ratio 3.5–4.5, sodium silicate 50 – 120 kg / m 3 , curing duration ≈ 35 days (diminishing returns beyond), constraining viable formulations from > 10 6 to ∼ 10 2 combinations. MCA-enhanced 3-day strength to > 40 MPa but required optimized chemical-temporal conditions for full effectiveness. An open-source graphical interface enables real-time prediction with integrated explainability, translating black-box models into actionable design tools. This framework advances geopolymer optimization from empirical practices toward transparent, mechanistically informed, and industrially deployable protocols for sustainable construction binders.
Neurodegenerative diseases (ND) are one of the most fatal diseases that affect the majority of individuals worldwide, among which Alzheimer’s disease (AD) and Parkinson’s disease (PD) are the most common. In vitro 2D monolayer cell cultures and in vivo transgenic animal models have been the primary tools for investigating mechanisms of neurodegenerative diseases. However, the ineffectiveness of these models in translating outcomes into human pathophysiology, necessitates innovative approaches to bridge the translational gap. In this review, we focus on the intricate pathogenic processes by which environmental toxicants and viral infections trigger neurodegeneration. The growing significance of three-dimensional (3D) brain organoids (BOs) derived from induced pluripotent stem cells (iPSCs) can be used as a groundbreaking platform for examining neurodegenerative pathways induced by exposure to environmental toxicants and viral infections. It also addressed how BO’s overcomes the fundamental limitations of traditional models, such as 2D cultures and animal models, thereby creating novel opportunities for the mechanistic study of multifactorial neurodegeneration and the development of therapeutic interventions.
The first stars, the chemically pristine Population III, likely played an important role in heating the intergalactic medium during the epoch of cosmic dawn. The very high effective temperatures (similar to 10(5) K) predicted for the most massive Population III stars could also give rise to tell-tale signatures in the emission-line spectra of early star clusters or small galaxies dominated by such stars. Important quantities in modelling their observational signatures include their photon production rates at ultraviolet energies at which photons are able to ionize hydrogen and helium, dissociate molecular hydrogen and cause Ly alpha heating. Here, we model the spectral energy distributions of Population III stars to explore how these key quantities are affected by the initial mass and rotation of Population III stars given a wide range of models for the evolution of these stars. Our results indicate that rotating Population III stars that evolve to effective temperatures similar to 2 x 10(5) K could potentially give rise to a very strong He ii 1640 angstrom emission line in the spectra from primordial star clusters, without requiring stellar masses of greater than or similar to 100 M-circle dot indicated by previous models for non-rotating Population III stars. The observable impact on 21-cm signatures from cosmic dawn and the epoch of reionization from our set of rotating stars that evolve to similar to 2 x 10(5) K is modest, except in case of high Population III star formation efficiencies which imprint potentially detectable features in the global 21-cm signal and 21-cm power spectrum.
Context. Orbiting matter misaligned with a spinning black hole undergoes Lense-Thirring precession due to the frame-dragging effect. This phenomenon is particularly relevant for type-C QPOs observed in the hard states of low-mass X-ray binaries. However, the accretion flow in these hard states is complex, consisting of a geometrically thick, hot corona surrounded by a geometrically thin, cold disk. Recent simulations demonstrate that, in such a truncated disk scenario, the precession of the inner, hot corona slows due to its interaction with the outer, cold disk. Aims. This paper aims to provide an analytical description of the precession of an inner (hot) torus in the presence of accretion torques exerted by the outer (cold) disk. Methods. Using the angular momentum conservation equation, we investigated the evolution of the torus angular momentum vector for various models of accretion torque. Results. We find that, in general, an accretion torque tilts the axis of precession away from the black hole spin axis. In all models, if the accretion torque is sufficiently strong, it can halt the precession; any perturbation from this stalled state causes the torus to precess around an axis that is misaligned with the black hole spin axis. Conclusions. The accretion torque exerted by the outer thin disk can cause precession around an axis that is neither aligned with the black hole spin axis nor perpendicular to the plane of the disk. This finding may have significant observational implications, as the jet direction, if aligned with the angular momentum axis of the torus, may no longer reliably indicate the black hole spin axis or the orientation of the outer accretion disk.
To enhance the performance of iterative algorithms, recent studies have emphasised the use of double inertial extrapolation steps. This work proposes a family of sufficient descent iterative algorithms incorporating a double inertial extrapolation strategy for solving nonlinear equations. Global convergence of the proposed methods is established under mild assumptions. Specifically, the global convergence is achieved without requiring Lipschitz continuity and assuming only generalised monotonicity, which is weaker than pseudo-monotonicity. In addition, under the local Lipschitz continuity assumption, we establish the asymptotic and non-asymptotic convergence rates in terms of iteration complexity. To the best of our knowledge, this work is the first to establish both asymptotic and non-asymptotic convergence rates for a derivative-free method under a double inertial framework. Furthermore, numerical experiments confirm the effectiveness and robustness of the proposed algorithms when compared with the recent inertial-based methods on standard test problems. Moreover, their applicability to regularised decentralised logistic regression and sparse signal restoration problems is demonstrated.
