Mean Shift is a flexible, non-parametric clustering algorithm that identifies dense regions in data through gradient ascent on a kernel density estimate. Its ability to detect arbitrarily shaped clusters without requiring prior knowledge of the number of clusters makes it widely applicable across diverse domains. However, its quadratic computational complexity restricts its use on large or high-dimensional datasets. Numerous acceleration techniques, collectively referred to as Fast Mean Shift strategies, have been developed to address this limitation while preserving clustering quality. This paper presents a systematic theoretical analysis of these strategies, focusing on their computational impact, pairwise combinability, and mapping onto distinct stages of the Mean Shift pipeline. Acceleration methods are categorized into seed reduction, neighborhood search acceleration, adaptive bandwidth selection, kernel approximation, and parallelization, with their algorithmic roles examined in detail. A pairwise compatibility matrix is proposed to characterize synergistic and conflicting interactions among strategies. Building on this analysis, we introduce a decision framework for selecting suitable acceleration strategies based on dataset characteristics and computational constraints. This framework, together with the taxonomy, combinability analysis, and scenario-based recommendations, establishes a rigorous foundation for understanding and systematically applying Fast Mean Shift methods.

The work studies the processes of contact melting between a mixture of refractory metal powders, carbon nanofibers and substrates made of steel plates of different grades, which arise after applying a current pulse to this composition from a contact welding machine. It has been established that in the proposed system, contact melting and sintering of individual components can occur selectively. This allows us to propose the use of such an effect for forming parts in additive technologies. The addition of carbon nanofibers ensures that contact melting occurs with lower energy costs and between certain components with the formation of materials of the desired structure, grain size, physical and mechanical properties. Experimental studies have shown that due to the rational selection of the composition of the charge, substrate materials and parameters of electrical pulses from the contact welding machine, it is possible to ensure selective directional melting of components and the formation of the material of the part with the desired properties. The introduction of refractory metals (Cr, W, Mo, V, etc.) and nanofibers into the composition of the charge allows us to ensure provide the required hardness, chemical resistance and strength of the material. A small content of metals inactive to carbon (Cu, Ag, Au, Ga, In, Ge, Sn, Pb, Bi and Sb) in the charge significantly improves the wetting and solubility in melts of carbon materials.

Carbohydrate metabolism disorders increase the risk of developing active tuberculosis and are associated with worse outcomes of tuberculosis treatment. 15% of tuberculosis cases worldwide are estimated to be related to diabetes mellitus. Altered lipid metabolism may influence susceptibility to tuberculosis. Studies have been published on the association between total cholesterol, low-density lipoprotein, and high-density lipoprotein levels and the risk of tuberculosis. Patients with tuberculosis have an increased incidence of cardiovascular and cerebrovascular diseases. Cardiovascular complications of tuberculosis are among the most common extrapulmonary manifestations of the disease.

This study presents the analysis and modeling of the thermal regime of a furnace lining at an industrial copper smelting facility using a combined approach based on neural network (NN) technologies and the finite element method (FEM). Experimental temperature data were collected from a laboratory setup equipped with three thermocouples (TP-2488/1 and TCRosemount 0065), with a sampling frequency of 1 Hz over a total duration of 5 hours, resulting in 18,000 measurement points. The measurement uncertainty of the thermocouples did not exceed ±1.5 °C. These data were used both for model development and for validating the numerical FEM simulations. A feedforward neural network was trained using 70% of the dataset, while 15% and 15% were used for validation and testing, respectively. The prediction error of the neural network remained within 3% with a 95% confidence interval of [2.6%, 3.4%]. The results show that the proposed hybrid approach improves temperature prediction accuracy and reduces static control error by 15% when combined with a proportional–integral controller. The methodology demonstrates significant potential for improving thermal process stability and reducing energy consumption in high-temperature metallurgical systems. © 2026 Institute of Advanced Engineering and Science. All rights reserved.

In the mining industry, when breaking rocks, oversized pieces are formed that exceed the dimensions of the technological equipment. To carry out secondary rock fragmentation in quarries, it is proposed to use an electromagnetic hammer. At the D.A. Kunayev Institute of Mining, a prototype of a medium class electromagnetic hammer with an impact energy of up to 6000 J has been developed. The use of the electromagnetic hammer for breaking oversized pieces in the mining industry will increase the efficiency, environmental friendliness, and automation of the technological processes of rock and ore fragmentation. To develop a competitive prototype of a rock-breaking hammer, it is necessary to justify its dynamic parameters. A microprocessor-based measurement system was developed to measure the dynamic parameters of the prototype of the electromagnetic hammer, followed by the calculation of the energy characteristics of the prototype using the software “Matlab.” A digital sensor “HC-SR04” and a microprocessor controller “Arduino Mega 2560” were used to measure the height of the moving part of the electromagnetic hammer. To eliminate measurement errors, calculate the dynamics of the plunger’s velocity, acceleration and impact energy values, and plot the dependency graphs, the “Matlab” program version R2021b was used. Experiments were conducted by supplying the power voltage to the upper coil with a current value of 120 amperes, while the lower coils received the power voltage with current values of 70, 100, and 120 amperes. The data obtained, after processing in the “Matlab” software, allowed for the construction of dynamic parameter graphs for the prototype. The analysis of the graphs revealed the design flaws of this prototype, the elimination of which will improve the technological characteristics of new developments. © 2025, National Academy of Sciences of the Republic of Kazakhstan. All rights reserved.

The object of the study is an IEEE 802.15.4 (2.4 GHz) wireless networked control system (WNCS) closing the loop over a wireless sensor network. Fading and interference increase packet loss and delay, reducing stability margins and control quality. The unresolved problem is the lack of a unified end-to-end (E2E) loss model that links PHY signal quality, multi-hop routing and medium access to closed-loop behavior and can be embedded into controller synthesis. An SINR-based channel model (path loss, lognormal shadowing, multipath fading) is mapped to BER and packet error probability; E2E loss for single-hop and multi-hop routes is obtained using Bernoulli and finite-state Markov (FSMC) processes. For verification, original packet traces are captured with an IEEE 802.15.4 sniffer/logger and stored before processing (timestamp, node identifier, sequence number, RSSI/LQI and delivery outcome) to compute PER, latency and burstiness and to parameterize the SINR-to-PER mapping and loss models. Simulations show that TDMA/TSCH achieves up to 40% lower loss than CSMA/CA, while E2E loss rises from 3% to 32% as hop count increases from 1 to 8. An MPC-based co-design jointly adapts transmit power, sampling period and retransmissions. Compared with a fixed-parameter LQR baseline, E2E PER is reduced from 4.45% to 3.66%, average delay from 0.20 s to 0.12 s, and integral absolute error by 50%. The gains are attributed to reduced contention under TDMA scheduling and predictor-driven MPC adaptation. The approach targets industrial monitoring and control with fixed sampling, slowly varying interference and static multi-hop topologies, where parameters can be identified offline and used for online MPC adaptation Copyright

As the volume of data continues to surge, researchers are confronted with the challenge of extracting meaningful insights from this wealth of information. Despite rapid advancements in Natural Language Processing (NLP) techniques in the AI industry, there remain gaps and opportunities for further exploration, particularly in the realm of data recombination techniques and methods. This paper proposes a novel text recombination method to facilitate the generation of recombined words from a given text. The process commences with a ’stanza’ model, which identifies and compiles Named Entity Recognitions (NERs) into a list. These NERs are then cross-referenced with Wikipedia pages to retrieve relevant information, thereby enhancing entity understanding and analysis. The ensuing step involves preprocessing the output text from the previous stage, generating a list of unique words while eliminating stop words. This preprocessing stage serves to remove noise and focus on meaningful words, laying the groundwork for more effective clustering. To enable clustering, we employ vector embeddings, representing words in a 2-dimensional space, rendering them suitable for clustering techniques. Notably, the proposed method further enhances results by re-clustering words after applying K-Means, thereby identifying the most fitting candidate words for recombination. Comparatively, this method outperforms large language models (LLMs) due to its incorporation of NER information, utilization of Wikipedia pages, and effective preprocessing techniques. Unlike LLMs, which operate as resource-intensive black boxes on static data, this method benefits from real-time information access and knowledge base updates. Furthermore, each stage of the process is visualized to control the progress correctly. Thus, due to the plots of word clusterization, the proposed text recombination approach showed positive results.

The article presents methods for calculating the production function based on complex-valued economics. The uniqueness of the work lies in the fact that the production function is presented as a criterion for managing and forecasting heat supply. Data from the boiler plants of the heat supply company in Taldykurgan are used as initial data. As a result of the analysis, it was decided to use the production function of the complex argument, which will increase the number of process indicators and conduct a more complete analysis of the production of thermal energy.

In the near future, AI may move beyond its role as a mere tool to function as a creative agent—potentially even as a virtual student or professor—capable of generating original artworks and contributing to research leadership. However, it remains unclear whether educational institutions are adequately prepared for such a rapid integration of AI into educational and research processes. This question becomes particularly relevant in the context of the rapid advancement of large language models (LLMs) and generative artificial intelligence (GenAI), given their potential to transform both the landscape of scientific research and educational methodologies. This study, therefore, examines how educational institutions are responding to the integration of AI into research and education. Specifically, we analyzed the policies and guidelines regulating the use of GenAI in both general universities and art-focused institutions, and conducted a strategic review of institutional approaches, along with a content analysis of selected curricula related to GenAI implementation. Based on the analysis, we posit that current GenAI policies in higher education are largely reactive, unevenly implemented across regions and disciplines, and often fail to address research-specific use cases and the distinct challenges faced by artfocused institutions. This finding aligns with recent studies showing that institutions tend to conform to external regulatory, normative, and mimetic pressures in their adoption of GenAI, often prioritizing legitimacy and compliance over proactive strategic vision (Singh, 2024). From an educational perspective, researchers further argued that Bloom’s Taxonomy requires revision to address the cognitive, affective, and metacognitive demands of AIassisted learning, underscoring the need for institutional policies that not only regulate GenAI but also foster critical thinking, ethical reasoning, and iterative learning processes in higher education (Gonsalves, 2024).

This article presents a mathematical model in the form of static equations of dependencies of input and output flows based on the equations of material and heat balance for the purposes of operational planning and control of the complex technological complex of Vanyukov melting (PV). Dynamic characteristics are presented for the purpose of controlling the thermal regime based on the technology of the developed melting process with blowing from below. As a result of the study, the developed mathematical model for controlling the smelting process when calculating the material flows of the charge will allow tracking changes in the thermal state of the smelting (by the copper content in the matte). This model can quite well describe the dynamics of the state of the process, both when establishing the impacts aimed at increasing the heating of the furnace, and at reducing its heating. Based on the equations, a computer model based on the dynamic programming method in the MATLAB software package has been developed. The scientific novelty lies in the fact that for the first time, the structure of a mathematical model has been developed that describes the processes occurring in the over-tuyere zone and the sludge zone of the smelting products. © 2023, Institute of Metallurgy and Ore Beneficiation JSC. All rights reserved.