Rajamangala University of Technology (Thai: มหาวิทยาลัยเทคโนโลยีราชมงคล), (RMUT), is one of the university systems in Thailand. It has nine universities providing undergraduate and graduate level education. It was elevated to university status in 2005. Before that it was known as Rajamangala Institute of Technology (สถาบันเทคโนโลยีราชมงคล).In September 2016, Prime Minister Prayut Chan-o-cha invoked Section 44 of the interim charter allowing him to form a special panel to take over administration of Rajamangala University of Technology Tawan-ok as it was judged to be incapable of administering itself.
Generative artificial intelligence (GenAI) is rapidly reshaping higher education, yet evidence remains limited on how students’ AI dependency relates to cognitive load and learning-related outcomes. Drawing on Technology Dependency Theory and Cognitive Load Theory, this study examines the curvilinear associations between AI dependency, intrinsic, extraneous, and germane cognitive load, and two outcomes: higher-order thinking skills (self-reported) and academic enjoyment. An online survey of 951 undergraduate students from eight universities in Sichuan, China, instructed participants to respond with reference to typical coursework in which they use the Chinese GenAI tool DeepSeek for study-related activities, rather than a one-off experimental task. Confirmatory factor analysis and structural equation modeling assessed measurement quality and estimated direct associations, while polynomial regression and the Two-Lines procedure tested nonlinearity.Results show a dependency curve. Extraneous load is lowest at moderate AI dependency. Germane load, higher-order thinking skills, and academic enjoyment also peak at moderate AI dependency. Intrinsic load decreases as more processing is handled by GenAI. The study shifts attention from usage amount to AI dependency. It shows outcomes are clearer when considering extraneous and germane load instead of assuming linear effects. The findings support structure-first guidance that reduces coordination and checking costs while maintaining generative engagement.
Probabilistic condition and safety maintenance strategies on concrete highway bridges with chloride-based deterioration are cost-prioritized, using a self-developed Monte Carlo platform. Practical probabilistic data of four maintenance types are gathered; Silane treatment (SL), Cathodic protection (CP), Minor concrete repair (CR), and Rebuild (RB). Six combined maintenance strategies (2–3 maintenance types) are proposed. From the study, the combined strategy of SL + RB is found cost-optimum based on 50-year present value of expected cumulative cost, nevertheless there is 55
Carotenoid biosynthesis in citrus is intricately regulated by developmental cues and environmental stimuli; however, elucidating these regulatory networks in planta remains challenging due to environmental variability. This study employed an in vitro culture system using juice sacs from ‘Siam Red Ruby’ and ‘Benimadoka’ pumelo cultivars to examine the influence of environmental factors on carotenoid metabolism under controlled conditions. Throughout a four-week culture period at 25 °C, the juice sacs exhibited cellular enlargement without callus formation, accompanied by significant changes in carotenoid content and gene expression. Treatment with blue LED light (470 nm peak), 5
Long-term time series forecasting (LTSF) is widely recognized as a central challenge in data mining and machine learning. LTSF has increasingly evolved into a benchmark-driven ”GAME,” where models are ranked, compared, and declared state-of-the-art based primarily on marginal reductions in aggregated pointwise error metrics such as MSE and MAE. Across a small set of canonical datasets and fixed forecasting horizons, progress is communicated through leaderboard-style tables in which lower numerical scores define success. In this GAME, what is measured becomes what is optimized, and incremental error reduction becomes the dominant currency of advancement. We argue that this metric-centric regime is not merely incomplete, but structurally misaligned with the broader objectives of forecasting. In real-world settings, forecasting often prioritizes preserving temporal structure, trend stability, seasonal coherence, robustness to regime shifts, and supporting downstream decision processes. Optimizing aggregate pointwise error does not necessarily imply modeling these structural properties. As a result, leaderboard improvement may increasingly reflect specialization in benchmark configurations rather than a deeper understanding of temporal dynamics. This paper revisits LTSF evaluation as a foundational question in data science: what does it mean to measure forecasting progress? We propose a multi-dimensional evaluation perspective that integrates statistical fidelity, structural coherence, and decision-level relevance. By challenging the current metric monoculture, we aim to redirect attention from winning benchmark tables toward advancing meaningful, context-aware forecasting.
High-manganese austenitic steels (e.g., Mn13) are critical for railway applications due to their exceptional work-hardening capacity, but their welding remains challenging due to carbide embrittlement and grain coarsening in the heat-affected zone (HAZ). This study proposes an integrated welding strategy combining in situ substrate cooling and multi-layer deposition during flux-cored arc welding (FCAW) to control thermal cycles, microstructure, and mechanical properties. The effects of four substrate cooling conditions—air (25 °C), water (20 °C), ice (0 °C), and ice + salt (−20 °C)—and one to three weld layers were systematically investigated. Results show that ice + salt cooling at −20 °C combined with three-layer deposition reduced the HAZ width by 56
