Evaluation Approach Research Articles

Enhancing practicality of deep learning for crop disease identification under field conditions: insights from model evaluation and crop-specific approaches.

Crop diseases can lead to significant yield losses and food shortages if not promptly identified and managed by farmers. With the advancements in convolutional neural networks (CNN) and the widespread availability of smartphones, automated and accurate identification of crop diseases has become feasible. However, although previous studies have achieved high accuracy (>95%) under laboratory conditions (Lab) using mixed data sets of multiple crops, these models often falter when deployed under field conditions (Field). In this study, we aimed to evaluate disease identification accuracy under Lab, Field, and Mixed (Lab and Field) conditions using an assembled data set encompassing 14 diseases of apple (Malus × domestica Borkh.), potato (Solanum tuberosum L.), and tomato (Solanum lycopersicum L.). In addition, we investigated the impact of model architectures, parameter sizes, and crop-specific models (CSMs) on accuracy, using DenseNets, ResNets, MobileNetV3, EfficientNet, and VGG Nets. Our results revealed a decrease in accuracy across all models from Lab (98.22%) to Mixed (91.76%) to Field (71.55%) conditions. Interestingly, disease classification accuracy showed minimal variation across model architectures and parameter sizes: Lab (97.61-98.76%), Mixed (90.76-92.31%), and Field (68.56-73.81%). Although CSMs were found to reduce inter-crop disease misclassifications, they also led to a slight increase in intra-crop misclassifications. Our findings underscore the importance of enriching data representation and volumes over employing new model architectures. Furthermore, the need for more field-specific images was highlighted. Ultimately, these insights contribute to the advancement of crop disease identification applications, facilitating their practical implementation in farmer's fields. © 2024 Society of Chemical Industry.

Multi objective optimization and evaluation approach of prefabricated component combination solutions using NSGA-II and simulated annealing optimized projection pursuit method

As a main carrier mode for the sustainable development of the construction industry in China, prefabricated building may lead to problems such as cost overruns, project delays, and waste of resources due to unreasonable selection of prefabricated components. Therefore, we quantitatively analyze the contribution rate of quality optimization of prefabricated components using QFD-SEM. Under the constraints of prefabrication rate, quality optimization contribution rate, and expected values of various sub-goals, we propose a multi-objective optimization method for prefabricated component combinations based on cost, duration, and carbon emissions. By using NSGA-II to solve the model, we can obtain a set of optimal Pareto solutions for prefabricated component combinations. Based on the optimal Pareto solution set, we establish a multi-objective evaluation model using simulated annealing optimization projection tracing method, and select the optimal prefabricated component combination solution according to the projected eigenvalues of the solutions. An empirical study is conducted using an eleven-story framed building in Shenzhen, Guangdong Province, China as a case study. The results show that: (1) Using this method, optimal solutions can be obtained in an unbounded solution space, with the optimal solution having advantages over both fully cast-in-place and fully prefabricated solutions. Compared to the fully cast-in-place solution, the duration and carbon emissions are reduced by 36.62% and 12.74% respectively, while compared to the fully prefabricated solution, costs are reduced by 4.15%. (2) There is a certain negative correlation between the cost of prefabricated component combinations and duration, carbon emissions, and quality optimization, while there is a certain positive correlation with the prefabrication rate. (3) The size of the optimal projection direction vector based on the optimization objectives indicates that carbon emissions have the greatest impact on the evaluation results of the solutions.

Elemental Profiling for the Detection of Food Mixtures: A Proof of Principle Study on the Detection of Mixed Walnut Origins Using Measured and Calculated Data.

Element profiling is a powerful tool for detecting fraud related to claims of geographical origin. However, these methods must be continuously developed, as mixtures of different origins in particular offer great potential for adulteration. This study is a proof of principle to determine whether elemental profiling is suitable for detecting mixtures of the same food but from different origins and whether calculated data from walnut mixtures could help to reduce the measurement burden. The calculated data used in this study were generated based on measurements of authentic, unadulterated samples. Five different classification models and three regression models were applied in five different evaluation approaches to detect adulteration or even distinguish between adulteration levels (10% to 90%). To validate the method, 270 mixtures of walnuts from different origins were analyzed using inductively coupled plasma mass spectrometry (ICP-MS). Depending on the evaluation approach, different characteristics were observed in mixtures when comparing the calculated and measured data. Based on the measured data, it was possible to detect admixtures with an accuracy of 100%, even at low levels of adulteration (20%), depending on the country. However, calculated data can only contribute to the detection of adulterated walnut samples in exceptional cases.

Multimodal Imaging Approach for Tumor Treatment Response Evaluation in the Era of Immunotherapy.

Immunotherapy is likely the most remarkable advancement in lung cancer treatment during the past decade. Although immunotherapy provides substantial benefits, their therapeutic responses differ from those of conventional chemotherapy and targeted therapy, and some patients present unique immunotherapy response patterns that cannot be judged under the current measurement standards. Therefore, the response monitoring of immunotherapy can be challenging, such as the differentiation between real response and pseudo-response. This review outlines the various tumor response patterns to immunotherapy and discusses methods for quantifying computed tomography (CT) and 18F-fluorodeoxyglucose positron emission tomography (PET) in the field of lung cancer. Emerging technologies in magnetic resonance imaging (MRI) and non-FDG PET tracers are also explored. With immunotherapy responses, the role for imaging is essential in both anatomical radiological responses (CT/MRI) and molecular changes (PET imaging). Multiple aspects must be considered when assessing treatment responses using CT and PET. Finally, we introduce multimodal approaches that integrate imaging and nonimaging data, and we discuss future directions for the assessment and prediction of lung cancer responses to immunotherapy.

Efficacy of daily versus intermittent oral iron supplementation for prevention of anaemia among pregnant women: a systematic review and meta-analysis

The World Health Organization recommends daily oral supplementation of iron for prevention of maternal anaemia. However, the adverse effects due to daily supplementation leads to poor compliance among pregnant women. Also, the mucosal block theory suggests that intermittent oral iron may be more efficient than daily iron with respect to optimum absorption. Our meta-analysis reviewed the existing clinical studies for the efficacy of daily versus intermittent oral iron supplementation. In this systematic review and meta-analysis [PROSPERO ID:CRD42024498180], we searched PubMed, Google Scholar, Scopus, Science Direct and Cochrane database for studies published from 1st January 1970 to 31st December, 2023. Studies comparing daily and intermittent iron supplementation in pregnant women were included. The median intermittent iron dose was 120mg/day and daily iron dose was 60mg/day. The primary outcome was endpoint haemoglobin levels after iron supplementation. The data was analysed using the 'meta' and 'metafor' packages in RStudio using random effects model. The heterogeneity, publication bias, risk of bias and certainty of evidence were assessed using I2 statistics, funnel plots, Cochrane Risk of Bias 2 (ROB2) tool, and the Grades of Recommendation, Assessment, Development and Evaluation (GRADE) approach respectively. Of 4615 search results, 26 studies (n=4365 participants) were included in this meta-analysis. There was no significant difference (p=0.18) between the endpoint mean haemoglobin levels of the daily versus intermittent oral iron groups (standardized mean difference (SMD): 0.51, 95% CI:-0.23 to 1.24, I2=97%, low certainty evidence) irrespective of baseline anaemic status. However, the endpoint ferritin levels were significantly higher in the daily supplementation group (SMD: 0.85, 95% CI: 0.15-1.54, p=0.02, I2=97%, low certainty evidence). The adjusted odds ratio for nausea, (adjusted odds ratio (OR) 3.56, 95% CI: 2.23-5.69, p<0.001, I2=9%, moderate certainty evidence), diarrhoea (adjusted OR 5.40, 95% CI: 1.90-15.33, p=0.002, I2=0%, low certainty evidence) and constipation (adjusted OR 1.95, 95% CI: 1.21-3.14, p=0.006, I2=0%, moderate certainty evidence) was significantly higher in daily oral iron supplementation group. Intermittent oral iron supplementation with a median dose of 120mg/day demonstrates comparable efficacy to daily oral iron supplementation median dose of 60mg/day in increasing haemoglobin levels among pregnant women with a significant reduction in adverse events. There was no funding for this study.

Penerapan Pelatihan Higiene Sanitasi Untuk Meningkatkan Pengetahuan Bagi Penjamah Makanan Nasi Boran di Kota Lamongan

This study aims to determine whether hygiene and sanitation training can improve knowledge for food handlers of boran rice in Lamongan City. This research method uses quantitative with an evaluative descriptive approach. Data collection techniques through Pre Test and Post Test questions. Research respondents were all food handlers of Boran rice in Lamongan city. a total of 20 people. with descriptive analysis techniques with the help of the SPSS (Statistical Package For Social Science) program. The results showed: by comparing the results of the pre-test and post-test, it can be seen that the training was quite effective in increasing participants' understanding and knowledge of hygiene and sanitation, because the average posttest answer of food handler respondents obtained a higher posttest score than when the pretest was conducted. Evidenced by the average Pre Test score was 66.75 and the average Post Test score was 73.25. Where the score is in the very good category. So it can be said that there is an effect of the application of sanitation hygiene training on increasing knowledge for food handlers of boran rice in Lamongan City.

Patient-Specific Factors Predicting Renal Denervation Response in Patients With Hypertension: A Systematic Review and Meta-Analysis.

The accurate selection of patients likely to respond to renal denervation (RDN) is crucial for optimizing treatment outcomes in patients with hypertension. This systematic review was designed to evaluate patient-specific factors predicting the RDN response. We focused on individuals with hypertension who underwent RDN. Patients were categorized based on their baseline characteristics. The primary outcome was blood pressure (BP) reduction after RDN. Both randomized controlled trials and nonrandomized studies were included. We assessed the risk of bias using corresponding tools and further employed the Grading of Recommendations Assessment, Development, and Evaluation approach to assess the overall quality of evidence. A total of 50 studies were ultimately included in this systematic review, among which 17 studies were for meta-analysis. Higher baseline heart rate and lower pulse wave velocity were shown to be associated with significant antihypertensive efficacy of RDN on 24-hour systolic BP reduction (weighted mean difference, -4.05 [95% CI, -7.33 to -0.77]; weighted mean difference, -7.20 [95% CI, -9.79 to -4.62], respectively). In addition, based on qualitative analysis, higher baseline BP, orthostatic hypertension, impaired baroreflex sensitivity, and several biomarkers are also reported to be associated with significant BP reduction after RDN. In patients with hypertension treated with the RDN, higher heart rate, and lower pulse wave velocity were associated with significant BP reduction after RDN. Other factors, including higher baseline BP, hypertensive patients with orthostatic hypertension, BP variability, impaired cardiac baroreflex sensitivity, and some biomarkers are also reported to be associated with a better BP response to RDN.

Can the Digital Economy Promote Sustainable Improvement in the Quality of Employment for Chinese Residents?—Moderated Mediation Effect Test Based on Innovation Environments

Employment significantly impacts the national economy and people’s livelihoods and affects millions of households. How to effectively and sustainably improve the quality of employment for the population has thus become a key issue facing China at present. In recent years, under the strategic background of “Digital China” and “Employment Priority”, the digital economy has brought about profound changes in the efficiency, dynamics, and distribution of social production. It affects the quality of employment by creating jobs, reshaping employment patterns, and improving labor quality, but its impact and transmission path are still unclear. This research employs a multi-dimensional evaluation approach to score the digital economy and employment quality at the provincial level in China, utilizing data from the country’s provincial panels between 2012 and 2022, and examines the feasibility and effect pathways of the digital economy in promoting sustainable improvement in the quality of residential employment. The empirical results provide ample evidence of the significant role played by the digital economy in the sustainable improvement of the employment quality of the inhabitants. Regions with more robust innovation environments tend to benefit more from this impact, and there are discernible regional variations in the impact. The upgrading of industrial structure mediates this process, and the influencing mechanism is regulated by the innovation environment—that is, the innovation environment exerts a facilitating influence on the process of industrial restructuring. The findings herein offer novel insights into the intrinsic mechanism of the digital economy in promoting sustainable improvement in the quality of residential employment.

Effectiveness of eHealth Interventions in Improving Medication Adherence Among Patients With Cardiovascular Disease: Systematic Review and Meta-Analysis.

Nonadherence to medication among patients with cardiovascular diseases undermines the desired therapeutic outcomes. eHealth interventions emerge as promising strategies to effectively tackle this issue. The aim of this study was to conduct a network meta-analysis (NMA) to compare and rank the efficacy of various eHealth interventions in improving medication adherence among patients with cardiovascular diseases (CVDs). A systematic search strategy was conducted in PubMed, Embase, Web of Science, Cochrane, China National Knowledge Infrastructure Library (CNKI), China Science and Technology Journal Database (Weipu), and WanFang databases to search for randomized controlled trials (RCTs) published from their inception on January 15, 2024. We carried out a frequentist NMA to compare the efficacy of various eHealth interventions. The quality of the literature was assessed using the risk of bias tool from the Cochrane Handbook (version 2.0), and extracted data were analyzed using Stata16.0 (StataCorp LLC) and RevMan5.4 software (Cochrane Collaboration). The certainty of evidence was evaluated using the Grading of Recommendations, Assessment, Development, and Evaluations (GRADE) approach. A total of 21 RCTs involving 3904 patients were enrolled. The NMA revealed that combined interventions (standardized mean difference [SMD] 0.89, 95% CI 0.22-1.57), telephone support (SMD 0.68, 95% CI 0.02-1.33), telemonitoring interventions (SMD 0.70, 95% CI 0.02-1.39), and mobile phone app interventions (SMD 0.65, 95% CI 0.01-1.30) were statistically superior to usual care. However, SMS compared to usual care showed no statistical difference. Notably, the combined intervention, with a surface under the cumulative ranking curve of 79.3%, appeared to be the most effective option for patients with CVDs. Regarding systolic blood pressure and diastolic blood pressure outcomes, the combined intervention also had the highest probability of being the best intervention. The research indicates that the combined intervention (SMS text messaging and telephone support) has the greatest likelihood of being the most effective eHealth intervention to improve medication adherence in patients with CVDs, followed by telemonitoring, telephone support, and app interventions. The results of these network meta-analyses can provide crucial evidence-based support for health care providers to enhance patients' medication adherence. Given the differences in the design and implementation of eHealth interventions, further large-scale, well-designed multicenter trials are needed. INPLASY 2023120063; https://inplasy.com/inplasy-2023-12-0063/.

Evaluating and Enhancing Artificial Intelligence Models for Predicting Student Learning Outcomes

Predicting student outcomes is an essential task and a central challenge among artificial intelligence-based personalised learning applications. Despite several studies exploring student performance prediction, there is a notable lack of comprehensive and comparative research that methodically evaluates and compares multiple machine learning models alongside deep learning architectures. In response, our research provides a comprehensive comparison to evaluate and improve ten different machine learning and deep learning models, either well-established or cutting-edge techniques, namely, random forest, decision tree, support vector machine, K-nearest neighbours classifier, logistic regression, linear regression, and state-of-the-art extreme gradient boosting (XGBoost), as well as a fully connected feed-forward neural network, a convolutional neural network, and a gradient-boosted neural network. We implemented and fine-tuned these models using Python 3.9.5. With a keen emphasis on prediction accuracy and model performance optimisation, we evaluate these methodologies across two benchmark public student datasets. We employ a dual evaluation approach, utilising both k-fold cross-validation and holdout methods, to comprehensively assess the models’ performance. Our research focuses primarily on predicting student outcomes in final examinations by determining their success or failure. Moreover, we explore the importance of feature selection using the ubiquitous Lasso for dimensionality reduction to improve model efficiency, prevent overfitting, and examine its impact on prediction accuracy for each model, both with and without Lasso. This study provides valuable guidance for selecting and deploying predictive models for tabular data classification like student outcome prediction, which seeks to utilise data-driven insights for personalised education.

Surface construction vibration impacts on underground tunnels: Approaches for using measurements and finite-element modeling to predict vibration in the frequency and time domain

Quantitative analyses of vibration impacts on underground tunnel structures were carried out on two projects. The first project was a new facility situated above an existing physics laboratory accelerator tunnel embedded in soil where the concern was vibration interference with sensitive research. Published and in-house data for construction equipment source levels were combined with measured transfer functions, between the ground surface and the tunnel, to estimate vibration transmitted to the tunnel in the frequency domain. This case study presents an approach for site-specific evaluations of construction vibration where spectral data are needed for evaluating vibration-sensitive equipment/receivers. The second project was a new facility situated above a storm relief tunnel embedded in rock where the concern was for potential vibration damage to the tunnel. A 3-D finite-element model of the rock and tunnel structure was developed to calculate the peak particle velocity and strain response of the tunnel walls in the time domain. This case study presents an approach for site-specific evaluations of construction vibration where peak velocity amplitudes and strain data are needed for evaluating structural damage potential.

A multi-criteria compromise ranking decision-making approach for analysis and evaluation of community-integrated energy service system

With the acceleration in industrialisation and urbanisation, building a community-integrated energy service system that includes renewable energy is an effective way to resolve energy shortages and reduce environmental pollution. Energy coupling can be realised using technology, such as electricity-to-gas conversion equipment, gas turbines, and gas boilers. And a carbon-trading mechanism can be introduced to build a community-integrated energy service system. First, the composition and characteristics of the community-integrated energy service system are analysed, and an index evaluation model for the community-integrated energy system is constructed considering the four influencing factors of cost, reliability, energy consumption, and environmental protection. The weight assigned to each indicator is determined based on the binary comparison weighting method and the improved criteria importance though intercrieria correlation method, and a multi-criteria compromise ranking evaluation system is established. A community was selected as an example of sensitivity and comparative analysis to verify the correctness and rationality of evaluating a community-integrated energy service system using the multi-criteria compromise ranking evaluation system based on the binary comparison weighting method and the improved criteria importance though intercrieria correlation method.

Toxic effects of trace metal(loid) mixtures on aquatic organisms

The co-occurrence of metal (loid)s in realistic aquatic environments necessitates the evaluation of their combined effects. However, the generality of the additive effect hypothesis is contentious, particularly due to metal(loid)-metal(loid) interactions. The absence of systematic evaluation approaches restricts our ability to draw overall conclusions and make reliable predictions. In this study, we reviewed 1473 effect sizes from 38 publications, and classified all responses into seven main categories (from molecular to individual levels) according to their toxicological significance. Our meta-analysis revealed that metal(loid) mixtures had significant effects on aquatic organisms (33 %, 95 % CI 28 %–39 %, P < 0.05), along with significant response heterogeneity (Qt = 690,319.62, P < 0.0001; I2 = 99.95 %). Concurrently, we developed a Random Forest machine learning model to predict adverse effects and identify key variables. These two methods demonstrated that the toxicity of metal(loid) mixtures is primarily linked to the choice of toxicity endpoints, and the characteristics of metal(loid) mixtures. Our findings underscore the potential of combining meta-analysis with machine learning, a more systematic approach, to enhance the understanding and prediction of the adverse effects of metal(loid) mixtures, and they offer guidance for risk assessment and policy-making in complex environmental scenarios.

Empathy Capacity-Building through a Community of Practice Approach: Exploring Perceived Impacts and Implications

Empathy can be a powerful catalyst for caring action toward wildlife and nature more broadly. Given the critical developmental period of early childhood, interventions that build the capacity of educators to support empathy development in young children are needed. We report on an evaluation of an empathy-focused Community of Practice (CoP) that engaged 15 regional early childhood professionals (including four zoo preschool staff) in the co-creation of empathy practices that are deepened through a ‘Two-Worlds’ (Indigenous and Western) approach. Ripple effect mapping was the evaluation approach used, as its participatory nature and emphasis on reflection were well aligned with the CoP approach. The results suggest CoP participation influenced teachers’ capacity for supporting empathy development, which rippled outward from teacher-level impacts to positive impacts on children, families, and beyond. We also offer reflections on the use of a CoP approach to build individual, organizational, and community capacity for supporting empathy, as well as on the conduciveness of ripple effect mapping for evaluating capacity-building professional learning in the context of empathy conservation.

A Systematic Review Supporting the Endocrine Society Clinical Practice Guidelines on Vitamin D.

Low vitamin D status is common and is associated with various common medical conditions. To support the development of the Endocrine Society's Clinical Practice Guideline on Vitamin D for the Prevention of Disease. We searched multiple databases for studies that addressed 14 clinical questions prioritized by the guideline panel. Of the 14 questions, 10 clinical questions assessed the effect of vitamin D vs no vitamin D in the general population throughout the lifespan, during pregnancy, and in adults with prediabetes; 1 question assessed dosing; and 3 questions addressed screening with serum 25-hydroxyvitamin D (25[OH]D). The Grading of Recommendations Assessment, Development and Evaluation approach was used to assess certainty of evidence. Electronic searches yielded 37 007 citations, from which we included 151 studies. In children and adolescents, low-certainty evidence suggested reduction in respiratory tract infections with empiric vitamin D. There was no significant effect on select outcomes in healthy adults aged 19 to 74 years with variable certainty of evidence. There was a very small reduction in mortality among adults older than 75 years with high certainty of evidence. In pregnant women, low-certainty evidence suggested possible benefit on various maternal, fetal, and neonatal outcomes. In adults with prediabetes, moderate certainty of evidence suggested reduction in the rate of progression to diabetes. Administration of high-dose intermittent vitamin D may increase falls, compared to lower-dose daily dosing. We did not identify trials on the benefits and harms of screening with serum 25(OH)D. The evidence summarized in this systematic review addresses the benefits and harms of vitamin D for the prevention of disease. The guideline panel considered additional information about individuals' and providers' values and preferences and other important decisional and contextual factors to develop clinical recommendations.

Pharmacological interventions for preventing upper gastrointestinal bleeding in people admitted to intensive care units: a network meta-analysis

ObjectivesTo assess the efficacy and safety of pharmacological interventions for preventing upper gastrointestinal (GI) bleeding in people admitted to intensive care units (ICUs).Design and settingSystematic review and frequentist network meta-analysis...

Limit hypersurface state of the art Gaidai multivariate risk evaluation approach for offshore Jacket

Current study presents state of the art approach to risk and reliability assessment for multivariate nonlinear dynamic systems. Specifically, novel hypersurface Gaidai risks evaluation methodology has been presented for evaluating offshore structural risks. Advocated approach being particularly suitable for offshore engineering multidimensional dynamic systems, possessing large number of critical components, that have either been physically observed/measured, or numerically modeled over a representative timelapse. Advocated Gaidai hypersurface structural risks evaluation methodology is applicable for a wide range of engineering and industrial systems. This study demonstrates that given in situ environmental conditions, it is well possible to assess accurately risks of system’s failures, damages or hazards, caused by excessive structural dynamics. Multivariate dynamic systems, possessing nonlinear cross-correlations between critical system’s components often present design challenges, when utilizing classic structural risks evaluation approaches, as those are mostly only univariate or bivariate. Dynamic offshore Jacket hot spot stresses have been utilized as an example in this reliability investigation. Modeling system’s excessive dynamics is challenging because of the non-stationarity of the system and the intricacy of fluid-structural dynamic interactions, arising from in situ wave-induced loads, influencing Jacket’s structural dynamics. Nonlinearities greatly affect structural dynamics, e.g., 2nd, 3rd, higher order effects within fluid-structural interactions. Methodology presented in this work offers a straightforward, effective, yet precise means of assessing the risks of failure or hazard for multivariate, non-stationary, non-linear dynamic offshore systems. For bivariate case, verification note has been added.

The life cycle of large language models in education: A framework for understanding sources of bias

AbstractLarge language models (LLMs) are increasingly adopted in educational contexts to provide personalized support to students and teachers. The unprecedented capacity of LLM‐based applications to understand and generate natural language can potentially improve instructional effectiveness and learning outcomes, but the integration of LLMs in education technology has renewed concerns over algorithmic bias, which may exacerbate educational inequalities. Building on prior work that mapped the traditional machine learning life cycle, we provide a framework of the LLM life cycle from the initial development of LLMs to customizing pre‐trained models for various applications in educational settings. We explain each step in the LLM life cycle and identify potential sources of bias that may arise in the context of education. We discuss why current measures of bias from traditional machine learning fail to transfer to LLM‐generated text (eg, tutoring conversations) because text encodings are high‐dimensional, there can be multiple correct responses, and tailoring responses may be pedagogically desirable rather than unfair. The proposed framework clarifies the complex nature of bias in LLM applications and provides practical guidance for their evaluation to promote educational equity. Practitioner notesWhat is already known about this topic The life cycle of traditional machine learning (ML) applications which focus on predicting labels is well understood. Biases are known to enter in traditional ML applications at various points in the life cycle, and methods to measure and mitigate these biases have been developed and tested. Large language models (LLMs) and other forms of generative artificial intelligence (GenAI) are increasingly adopted in education technologies (EdTech), but current evaluation approaches are not specific to the domain of education. What this paper adds A holistic perspective of the LLM life cycle with domain‐specific examples in education to highlight opportunities and challenges for incorporating natural language understanding (NLU) and natural language generation (NLG) into EdTech. Potential sources of bias are identified in each step of the LLM life cycle and discussed in the context of education. A framework for understanding where to expect potential harms of LLMs for students, teachers, and other users of GenAI technology in education, which can guide approaches to bias measurement and mitigation. Implications for practice and/or policy Education practitioners and policymakers should be aware that biases can originate from a multitude of steps in the LLM life cycle, and the life cycle perspective offers them a heuristic for asking technology developers to explain each step to assess the risk of bias. Measuring the biases of systems that use LLMs in education is more complex than with traditional ML, in large part because the evaluation of natural language generation is highly context‐dependent (eg, what counts as good feedback on an assignment varies). EdTech developers can play an important role in collecting and curating datasets for the evaluation and benchmarking of LLM applications moving forward.

Матрица бизнес-выгод как инструмент исследования потенциала цифровизации на примере мини-завода для производства невзрывчатых компонентов эмульсионных взрывчатых веществ

Modern business solutions set a trend towards integrating digital technologies into all areas of corporate organization. Digitalization plays a key role in modern business, enabling companies to increase the process efficiency, reduce costs, and improve product quality and maintenance efficiency. It also helps to be competitive in the marketplace and adapt to changing customer demands. In order to maximize the effects of implementing digitalization, the transformation needs to encompass not only the local business processes and business models, but also the corporate culture. The study analyses the process of exploring the digitalization potential for a modular mobile plant (mini-plant) that produces non-explosive components of emulsion explosives developed by AZOTTECH. The paper identifies the key points necessary for successful implementation of the digitalization process at the mini-plant, which can serve as a basis for designing a digital twin implementation strategy. It also describes application of a modern approach in performance evaluation using the matrix of business benefits.

A model-based approach for formal verification and performance evaluation of energy harvesting architectures in IoT systems: A case study of a long-term healthcare application

Energy harvesting plays a significant role in the Internet of Things (IoT). Indeed, although numerous approaches exist to limit the system’s power consumption, the energy provided by the battery remains constrained, thereby limiting the system’s lifetime. Energy harvesting represents an interesting technique that allows a set of devices in an IoT architecture to operate for a potentially infinite time without the need for battery replacement or recharge. This work presents a formal modeling framework for the performance evaluation of energy harvesting architectures and strategies in IoT systems. We present a model-based approach using UPPAAL to model and analyze IoT device lifetimes and capture the energy-related behavior of nodes and various energy harvesters. Furthermore, the model is calibrated using measurements acquired from real-life IoT applications to demonstrate the effectiveness of the proposed model and its ability to investigate various energy-related aspects.