Accurate univariate forecasting remains a pressing need in real-world systems, such as energy markets, hydrology, retail demand, and IoT monitoring, where signals are often intermittent and horizons span both short- and long-term. While transformers and Mixture-of-Experts (MoE) architectures are increasingly favored for time-series forecasting, a key gap persists: MoE models typically require complicated training with both the main forecasting loss and auxiliary load-balancing losses, along with careful routing/temperature tuning, which hinders practical adoption. In this paper, we propose a model architecture that simplifies the training process for univariate time series forecasting and effectively addresses both long- and short-term horizons, including intermittent patterns. Our approach combines sparse MoE computation with a novel attention-inspired gating mechanism that replaces the traditional one-layer softmax router. Through extensive empirical evaluation, we demonstrate that our gating design naturally promotes balanced expert utilization and achieves superior predictive accuracy without requiring the auxiliary load-balancing losses typically used in classical MoE implementations. The model achieves better performance while utilizing only a fraction of the parameters required by state-of-the-art transformer models, such as PatchTST. Furthermore, experiments across diverse datasets confirm that our MoE architecture with the proposed gating mechanism is more computationally efficient than LSTM for both long- and short-term forecasting, enabling cost-effective inference. These results highlight the potential of our approach for practical time-series forecasting applications where both accuracy and computational efficiency are critical.
We study the stress-strain state of an inhomogeneous orthotropic circular closed cylindrical shell subjected to initial axial loading and local heating by convective heat exchange. The material properties of the shell exhibit exponential variations across the thickness. In our investigation, we use a mathematical model of the shear theory of Timoshenko-type inhomogeneous shells. The two-dimensional heat-conduction equation is deduced under the condition of linear dependence of temperature on the thickness coordinate. By using the Laplace integral transform with respect to time and the finite Fourier transform with respect to spatial coordinates, we construct analytic solutions of the nonstationary problem of thermoelasticity for a finite hinged circular cylindrical shell. We perform the numerical analysis of the influence of the initial plane stressed state, the inhomogeneity parameter, and other thermophysical parameters on the displacements and stresses formed in the shell.
We present a combined ML/NLP (Machine Learning, Natural Language Processing) pipeline for protecting cloud-based APIs (Application Programming Interfaces), which works both at the level of individual HTTP (Hypertext Transfer Protocol) requests and at the access log file reading mode, linking explicitly technical anomalies with business risks. The system processes each event/access log through parallel numerical and textual branches: a set of anomaly detectors trained on traffic engineering characteristics and a hybrid NLP stack that combines rules, TF-IDF (Term Frequency-Inverse Document Frequency), and character-level models trained on enriched security datasets. Their results are integrated using a risk-aware policy that takes into account endpoint type, data sensitivity, exposure, and authentication status, and creates a discrete risk level with human-readable explanations and recommended SOC (Security Operations Center) actions. We implement this design as a containerized microservice pipeline (input, preprocessing, ML, NLP, merging, alerting, and retraining services), orchestrated using Docker Compose and instrumented using OpenSearch Dashboards. Experiments with OWASP-like (Open Worldwide Application Security Project) attack scenarios show a high detection rate for injections, SSRF (Server-Side Request Forgery), Data Exposure, and Business Logic Abuse, while the processing time for each request remains within real-time limits even in sequential testing mode. Thus, the pipeline bridges the gap between ML/NLP research for security and practical API protection channels that can evolve over time through feedback and retraining.
The undoped ZnAl2O4 and MgAl2O4 spinels are investigated by the X-ray diffraction and the steady-state and time-resolved luminescence spectroscopy methods in the 6-520 K temperature range. A new violet emission band with the maximum around 3.1 eV is observed in the luminescence spectrum of these materials. The excitation spectrum of the approximate to 3.1 eV emission consists of three bands located at about 3.65 eV, 4.95 eV, and 5.85 eV. The Stokes shift of the emission band is about 0.55 V. The luminescence decay time is short (a few nanoseconds) and independent of temperature in the investigated temperature range. The emission with such characteristics does not belong to any known impurity or intrinsic luminescence center in the spinels, but is similar to the F+ luminescence observed in many oxide materials. The suggestion is made that the violet emission arises from the center of the F+-type, which can be considered as the F+ center (an oxygen vacancy with one trapped electron) associated with an impurity or intrinsic crystal lattice defect for the excess volume and/or charge compensation. Possible structure and formation mechanisms of the F+-type centers in the investigated ZnAl2O4 and MgAl2O4 spinels are discussed.
We study the eigenfrequencies of transverse vibrations of a circular in-plane preheated plate with elastically restrained edge. We formulate the corresponding axisymmetric problem of thermoelasticity by using the theory of plates and construct its analytic solution with regard for the heat exchange (according to the Newton law) from the surfaces and the elastic compliance on the outer contour of the plate. The frequency equations of transverse vibrations are constructed for a plate prestressed as a result of uniform heating of the plate under the boundary conditions of elastically restrained edge. The influence of temperature and the elastic characteristics of restraint on the vibration characteristics of the round plate is analyzed.
