Search

This work presents a detailed case study on using Generative AI (GenAI) to develop AI surrogates for simulation models in fusion energy research. The scope includes the methodology, implementation, and results of using GenAI to assist in model development and optimization, comparing these results with previous manually developed models.

This study investigated (a) differences between males and females for changes in serum, tear, and urine osmolality, hematocrit, and urine specific gravity following acute passive dehydration and (b) assessed the reliability of these biomarkers separately for each sex. Fifteen males (age: 26.3 ± 3.5 years, body mass: 76 ± 7 kg) and 15 females (age: 28.8 ± 6.4 years, body mass: 63 ± 7 kg) completed a sauna protocol twice (5–28 days apart), aiming for 4% body mass loss (BML). Urine, blood, and tear markers were collected pre- and postdehydration, and change scores were calculated. Male BML was significantly greater than that of females in Trial 1 (3.53% ± 0.55% vs. 2.53% ± 0.43%, p <.001) and Trial 2 (3.36% ± 0.66% vs. 2.53% ± 0.44%, p =.01). Despite significant differences in BML, change in hematocrit was the only change marker that displayed a significant difference in Trial 1 (males: 3% ± 1%, females: 2% ± 1%, p =.004) and Trial 2 (males: 3% ± 1%, females: 1% ± 1%, p =.008). Regression analysis showed a significant effect for sex (male) predicting change in hematocrit (β = 0.8, p =.032) and change in serum osmolality (β = −3.3, p =.005) when controlling for BML but not for urinary or tear measures. The intraclass correlation coefficients for females (ICC 2, 1) were highest for change in urine specific gravity (ICC =.62, p =.006) and lowest for change in tear osmolarity (ICC = −.14, p =.689), whereas for males, it was posthematocrit (ICC =.65, p =.003) and post tear osmolarity (ICC =.18, p =.256). Generally, biomarkers showed lower test–retest reliability in males compared with females but, overall, were classified as poor–moderate in both sexes. These findings suggest that the response and reliability of hydration biomarkers are sex specific and highlight the importance of accounting for BML differences. [ABSTRACT FROM AUTHOR]

It is not currently possible to directly and noninvasively measure in vivo patellofemoral joint contact force during dynamic movement; therefore, indirect methods are required. Simple models may be inaccurate because patellofemoral contact forces vary for the same knee flexion angle, and the patellofemoral joint has substantial out-of-plane motion. More sophisticated models use 3-dimensional kinematics and kinetics coupled to a subject-specific anatomical model to predict contact forces; however, these models are time consuming and expensive. We applied a principal component analysis prediction and regression method to predict patellofemoral joint contact forces derived from a robust musculoskeletal model using exclusively optical motion capture kinematics (external approach), and with both patellofemoral and optical motion capture kinematics (internal approach). We tested this on a heterogeneous population of asymptomatic subjects (n = 8) during ground-level walking (n = 12). We developed equations that successfully capture subject-specific gait characteristics with the internal approach outperforming the external. These approaches were compared with a knee-flexion based model in literature (Brechter model). Both outperformed the Brechter model in interquartile range, limits of agreement, and the coefficient of determination. The equations generated by these approaches are less computationally demanding than a musculoskeletal model and may act as an effective tool in future rapid gait analysis and biofeedback applications. [ABSTRACT FROM AUTHOR]

Previous work has shown that estimates of climate sensitivity vary over time in response to abrupt CO2 forcing in climate model simulations. The energy fluxes that drive warming in response to increasing CO2 also influence precipitation, which prompts the question: Does the precipitation response therefore also vary over time? We investigate by examining the response of precipitation to warming forced by greenhouse gases—the hydrological sensitivity—in a set of millennial‐length climate simulations with multiple climate models, Long Run Model Intercomparison Project (LongRunMIP). We compare hydrological sensitivity calculated from three different timescales of the simulations: years 1–20, 21–150, and 151–1000. We show that the hydrological sensitivity lacks a consistent dependence on timescale, in contrast to climate sensitivity. Decomposition of the surface energy budget reveals that the relative muting of the multi‐model mean hydrological sensitivity is driven by surface downwelling shortwave flux. Plain Language Summary: Previous work has shown that when carbon dioxide concentrations are abruptly quadrupled, the response of the global energy imbalance between the Earth and space is a fast change for the first couple of decades, followed by a slower response as time goes on. Precipitation is also an important part of the energy budget, so perhaps its response to increasing carbon dioxide might depend on timescale as well. This study investigates using a recent set of very long model simulations. We find that there is less of a dependence for precipitation than for the global energy imbalance. Because the effect is smaller, it is also more subject to the details of the statistical analysis, so particular attention is paid there. Key Points: The trend across timescales of hydrological sensitivity is less consistent and has a smaller magnitude compared to climate sensitivityDownwelling surface shortwave flux is the component of the energy budget that contributes most consistently to this difference [ABSTRACT FROM AUTHOR]

This research explores the design of an infrared (IR) photodetector using mercury cadmium telluride (Hg1–xCdxTe). It proposes two- and three-dimensional homojunction models based on p+-Hg0.7783Cd0.2217Te/n–-Hg0.7783Cd0.2217Te, focusing on applications in the long-wavelength infrared range. The photodetector's performance is analyzed using Silvaco ATLAS TCAD software and compared with analytical calculations based on drift-diffusion, tunneling, and Chu's approximation techniques. Optimized for operation at 10.6 μm wavelength under liquid nitrogen temperature, the proposed photodetector demonstrates promising optoelectronic characteristics including the dark current density of 0.20 mA/cm2, photocurrent density of 4.98 A/cm2, and photocurrent density-to-dark current density ratio of 2.46 × 104, a 3-dB cut-off frequency of 104 GHz, a rise time of 0.8 ps, quantum efficiency of 58.30 %, peak photocurrent responsivity of 4.98 A/W, specific detectivity of 3.96 × 1011 cmHz1/2/W, and noise equivalent power of 2.52 × 10–16 W/Hz1/2 indicating its potential for low-noise, high-frequency and fast-switching applications. The study also incorporates machine learning regression models to validate simulation results and provide a predictive framework for performance optimization, evaluating these models using various statistical metrics. This comprehensive approach demonstrates the synergy between advanced materials science and computational techniques in developing next-generation optoelectronic devices. By combining theoretical modeling, simulation, and machine learning, the research highlights the potential to accelerate progress in IR detection technology and enhance device performance and efficiency. This multidisciplinary methodology could serve as a model for future studies in optoelectronics, illustrating how advanced materials and computational methods can be utilized to enhance device capabilities. [ABSTRACT FROM AUTHOR]

A novel approach to the construction of absolutely continuous distributions over the unit interval is proposed. Considering two absolutely continuous random variables with positive support, this method conditions on their convolution to generate a new random variable in the unit interval. This approach is demonstrated using some popular choices of positive random variables, such as the exponential, Lindley, and gamma. Some existing distributions, like the uniform, the beta, and the Kummer-beta, are formulated with this method. Several new structures of density functions having potential for future applications in real-life problems are also provided. One of the new distributions, namely the LCG, is considered for detailed study along with a related distribution, namely the GCL. The moments, hazard rate, cumulative distribution function, stress-strength reliability, random sample generation using the quantile function, method of moments along with maximum likelihood estimation, and regression modeling are discussed for both the distributions. Real-life applications of the proposed models and the corresponding regression models show promising results. [ABSTRACT FROM AUTHOR]

We investigate the influence of spin and impurity on the density of states of SiC nanotubes employing Density Functional Theory (DFT) and a Machine Learning (ML) based framework. Our study investigates the electronic structures and magnetic properties of various SiC nanotube configurations, including wurtzite, Co-doped, and undoped single-wall (6,0) chiral nanotubes, employing both DFT and pseudopotential approaches. The calculated energy band gap values for SiC bulk structures, nanotubes, and doped systems, retaining local density and local spin density approximations with the Hubbard U method, exhibit distinct characteristics. While undoped SiC systems remain nonmagnetic whereas Co-doped SiC systems show magnetic properties, with a total magnetic moment of around ~ 1.9 µB. Our first-principles calculations indicate that Co-doped SiC nanotubes induce magnetism, however the total energy calculations revealed satisfactory stability for the ferromagnetic phase. Validation against DFT data demonstrates that our model achieves approximately 91.5% accuracy for predicting the density of states for quantum-confined SiC nanotube structures and also showcasing significant potential for further electronic properties calculations in this domain. Integrating ML algorithms with DFT-based approach presents an efficient algorithm for predicting total density of states in quantum-confined nanoscale structures. The fine tree regression algorithm shows highly efficient and effective prediction for density of states calculations. [ABSTRACT FROM AUTHOR]

Background: Organizational turnover exacerbates the shortage of nurses in the global workforce. However, no study has yet explored how deep structure usage—nurses' integration of electronic patient records into nursing practice delivery—reduces their turnover intention and moderates the impact of affective, continuance, and normative professional commitment on their turnover intention. Aims: To ascertain (1) the linkage between the deep structure usage of electronic patient records and nurses' organizational turnover intention and (2) the moderating role of deep structure usage on the associations between elements of commitment (affective, continuance, and normative) and turnover intention. Methods: Using a cross‐sectional survey and proportionate random sampling by ward unit, we collected data from 417 full‐time nurses via a self‐administered questionnaire. We performed hierarchical regression analyses to test the study hypotheses. Results: Deep structure usage was not directly related to organizational turnover intention (β = −0.07, p = 0.06). However, the results suggested that deep structure usage may enhance the effect of high affective commitment on nurses' organizational turnover intention (β = −0.09, p = 0.04), while potentially mitigating the effect of low continuance commitment on organizational turnover intention (β = 0.10, p = 0.01). Conclusions: Deep structure usage of electronic patient records helps to ease nurses' workload and facilitates their retention, which is particularly due to their affective commitment (attachment) but not their continuance commitment (switching costs). Implications for Nursing Management: Nursing management may advise hospital management that medical records systems need to be improved and fully embedded for nursing care delivery, as a more in‐depth use of these systems can help to retain nurses. [ABSTRACT FROM AUTHOR]

BACKGROUND RESULTS CONCLUSION Radix Bupleuri has been widely used for its plentiful pharmacological effects. But it is hard to evaluate their safety and efficacy because the concentrations of components are tightly affected by the surrounding environment. Thus, Radix Bupleuri samples from different regions and varieties were collected. Based on the experimental and computational Raman spectrum, machine learning is emphasized for certain obscured characteristics; for example, linear discriminant analysis (LDA), support vector machine (SVM), eXtreme gradient boosting (XGBoost) and light gradient boosting machine (LightGBM).After dimension reduction by LDA, models of SVM, XGBoost and LightGBM were trained for classification and regression prediction of Bupleurum production regions. Support vector classifiers achieved the best accuracy of 98% and an F1 score above 0.96 on the test set. Support vector regression has a good fitting performance with an R2 score above 0.90 and a relatively low mean square error. However, complex models were prone to overfitting, resulting in poor generalization ability.Among these machine learning models, the typical LDA‐SVM models, consistent with the high‐performance liquid chromatography results, demonstrate great performance and stability. We envision that this rapid classification and regression technique can be extended to predictions for other herbs. © 2024 Society of Chemical Industry. [ABSTRACT FROM AUTHOR]

Introduction: Animal rabies is a viral disease that poses a significant threat to domestic and wild animal populations, with devastating consequences for animal health and human life. Understanding and assessing the risk factors associated with the transmission and persistence of the rabies virus in wild and domestic animal populations is crucial for developing effective strategies to control and mitigate cases. Studies of the spatial and temporal distribution of rabies cases in the Nizhny Novgorod region during 2012-2022 provided epidemiological evidence of the circulation of infection between animals in the presence of vaccination. Among the wild animals in the area, red foxes play a major role in the spread of rabies, accounting for 96.4% of all wild animal cases. Methods: We used spatiotemporal cluster analysis and a negative binomial regression algorithm to study the relationships between animal rabies burden by municipality and a series of environmental and sociodemographic factors. Results: The spatiotemporal cluster analysis suggests the concentration of wild animal rabies cases in the areas of high fox population density and insufficient vaccination rates. The regression models showed satisfactory performance in explaining the observed distribution of rabies in different animals (R 2 = 0.71, 0.76, and 0.79 in the models for wild, domestic and all animals respectively), with rabies vaccination coverage and fox population density being among the main risk factors. Conclusion: We believe that this study can provide valuable information for a better understanding of the geographical and temporal patterns of rabies distribution in different animal species, and will provide a basis for the development of density-dependent planning of vaccination campaigns. [ABSTRACT FROM AUTHOR]

A little less than half of the students of higher Ed institutions in the US graduate in four years, and only around 60% finish in six years. Retention rates are also less than ideal. Colleges have been experimenting with a variety of programs and policies to address this issue, especially less selective institutions whose rates are significantly lower. In this paper, we evaluate a student success and retention program called Early Alert that was implemented at a public state university in the US with a medium-to-large student body. Our dataset contains several years' worth of information on students' socio-demographic characteristics, class standing and average grades (GPAs), as well as their midterm and final grades in undergraduate courses. We employ several causal inference techniques developed for observational studies and elicit negative average treatment effects on the treated (ATT). Since it is conceivable that unobserved confounders are the real drivers of our empirical results, not the treatment, we carry out two different types of sensitivity analyses. Together with our treatment effect estimations, they lead us to the main conclusion that Early Alert does not improve student performance, at least not in the short run (as measured by course performance), and likely has a negligible impact. [ABSTRACT FROM AUTHOR]

Despite significant progress, a comprehensive analysis of Machine learning algorithms for detecting delamination and crack damage in composite laminate plates is noticeably absent in the literature. This study aims to bridge this gap by meticulously help spectrum of six regression algorithms. By utilizing simulated displacement-time responses from a 16-layer laminated composite plate structure subjected to a random force, a robust dataset is constructed to evaluate the efficacy of these machine learning paradigms. Performance assessments rely on critical metrics: coefficient of determination (R2), mean absolute error (MAE), root mean squared error (RMSE), and mean absolute percentage error (MAPE). [ABSTRACT FROM AUTHOR]

This article sheds light on the relationship between compositional data (CoDa) regression models and multiplicative competitive interaction (MCI) models, which are two approaches for modeling shares. We demonstrate that MCI models are particular cases of CoDa models with a total and that a reparameterization links both. Recognizing this relation offers mutual benefits for the CoDa and MCI literature, each with its own rich tradition. The CoDa tradition, with its rigorous mathematical foundation, provides additional theoretical guarantees and mathematical tools that we apply to improve the estimation of MCI models. Simultaneously, the MCI model emerged from almost a century-long tradition in marketing research that may enrich the CoDa literature. One aspect is the grounding of the MCI specification in assumptions on the behavior of individuals. From this basis, the MCI tradition also provides credible justifications for heteroskedastic error structures – an idea we develop further and that is relevant to many CoDa models beyond the marketing context. Additionally, MCI models have always been interpreted in terms of elasticities, a method that has only recently emerged in CoDa. Regarding this interpretation, the CoDa perspective leads to a decomposition of the influence of the explanatory variables into contributions from relative and absolute information. [ABSTRACT FROM AUTHOR]

In behavioral and social research, questionnaires are an important assessment tool, through which individuals can be categorized according to how they classify themselves in respect to a personal trait. One example is the Major Depression Inventory (MDI), which is widely used for the assessment of depression. It can also be used as a depression severity scale, with scores ranging from 0 to 50 constructed considering the same weight for each item in the MDI. However, the dependence among the items of the questionnaire suggests that a score with better properties could be obtained through factor models, which besides allowing to reduce the dimensionality of multivariate data, provides the estimation of common factors and factor loadings that often have an interesting theoretical interpretation. Additionally, auxiliary information could be available and, the effect of these variables in the latent factor could be estimated and provide interesting results. Thus, the main aim of this paper is to propose a factor model for ordered categorical data which incorporates auxiliary variables to explain the latent factors. The proposed model provides an alternative score to MDI based on the estimated latent factors that takes the uncertainty in the data and auxiliary information into account. [ABSTRACT FROM AUTHOR]

Composite minimization is a powerful framework in large-scale convex optimization, based on decoupling of the objective function into terms with structurally different properties and allowing for more flexible algorithmic design. We introduce a new algorithmic framework for complementary composite minimization, where the objective function decouples into a (weakly) smooth and a uniformly convex term. This particular form of decoupling is pervasive in statistics and machine learning, due to its link to regularization. The main contributions of our work are summarized as follows. First, we introduce the problem of complementary composite minimization in general normed spaces; second, we provide a unified accelerated algorithmic framework to address broad classes of complementary composite minimization problems; and third, we prove that the algorithms resulting from our framework are near-optimal in most of the standard optimization settings. Additionally, we show that our algorithmic framework can be used to address the problem of making the gradients small in general normed spaces. As a concrete example, we obtain a nearly-optimal method for the standard ℓ 1 setup (small gradients in the ℓ ∞ norm), essentially matching the bound of Nesterov (Optima Math Optim Soc Newsl 88:10–11, 2012) that was previously known only for the Euclidean setup. Finally, we show that our composite methods are broadly applicable to a number of regression and other classes of optimization problems, where regularization plays a key role. Our methods lead to complexity bounds that are either new or match the best existing ones. [ABSTRACT FROM AUTHOR]

Purpose: This study is to investigate whether angiotensin type 1 receptor blockers (ARBs) or angiotensin-converting enzyme inhibitors (ACEis) can regress coronary artery aneurysm (CAA) in patients with Kawasaki disease (KD). Methods: This multicenter, prospective, observational study was conducted at 53 institutions throughout Japan. We enrolled patients who were diagnosed with KD after January 2015 and had a medium or large CAA (maximum luminal diameter ≥ 4 mm or z score ≥ + 5) 30 days or later after KD onset. Results: Of the 209 patients, 47 (22%) were taking ARBs/ ACEis. Compared with those in the non-ARB/ACEi group, the baseline CAA diameter was significantly greater (6.7 mm vs. 5.5 mm, p < 0.01), and bilateral CAA (70% vs. 59%, p = 0.01) and giant CAA (32% vs. 20%, p = 0.08) were more frequently observed in the ARB/ACEi group. Although the overall regression rates did not differ between the groups (67% vs. 65%), the regression rates of giant CAA were approximately 1.6 times greater in the ARB/ACEi group than in the non-ARB/ACEi group (36% vs. 23%). Multivariate Cox regression analysis after adjustment for other clinical variables suggested that ARBs/ACEis may be a factor in CAA regression (hazard ratio [HR]: 1.5, 95% confidence interval [CI]: 0.91–2.46). Conclusions: Although ARBs/ ACEis were used more frequently in patients with severe CAA, these patients had similar CAA regression rates to patients not taking ARBs/ACEis. ARBs/ACEis may be beneficial agents aimed at inducing CAA regression in KD patients. What is Known: • Large CAAs are less likely to regress and are always at risk of life-threatening cardiac events. • Moderate CAA, age less than 1 year, and female sex have been reported to be factors that promote the regression of CAA. What is New: • Although ARBs/ACEis were used more frequently in patients with severe CAA, these patients had a similar rate of CAA regression to patients who did not take ARBs/ACEis. • The regression rates of giant CAA were approximately 1.6 times greater in the ARB/ACEi group than in the non-ARB/ACEi group. [ABSTRACT FROM AUTHOR]

We obtain rates of contraction of posterior distributions in inverse problems with discrete observations. In a general setting of smoothness scales we derive abstract results for general priors, with contraction rates determined by discrete Galerkin approximation. The rate depends on the amount of prior concentration near the true function and the prior mass of functions with inferior Galerkin approximation. We apply the general result to non-conjugate series priors, showing that these priors give near optimal and adaptive recovery in some generality, Gaussian priors, and mixtures of Gaussian priors, where the latter are also shown to be near optimal and adaptive. [ABSTRACT FROM AUTHOR]

Deep learning is revolutionizing various scientific fields, with medical applications at the forefront. One key focus is automating image segmentation, a process crucial in many clinical services. However, medical images are often ambiguous and challenging even for experts. To address this, reliable models need to quantify their uncertainty, allowing physicians to understand the model's confidence in its segmentation. This paper explores how the performance and uncertainty of a model are influenced by the number of annotations per input sample. We examine the effects of both single and multiple manual annotations on various deep learning architectures. To tackle this question, we employ three widely recognized deep learning architectures and evaluate them across four publicly available datasets. Furthermore, we explore the effects of dropout rates on Monte Carlo models by examining uncertainty models with dropout rates of 20%, 40%, 60%, and 80%. Subsequently, we evaluate the models using various measurement metrics. The findings reveal that the influence of multiple annotations varies significantly depending on the datasets. Additionally, we observe that the dropout rate has minimal or no impact on the model's performance unless there is a substantial loss of training data, primarily evident in the 80% dropout rate scenario. [ABSTRACT FROM AUTHOR]

Brain age serves as a vital biomarker for detecting neurological ailments like Alzheimer's disease (AD) and Parkinson's disease (PD). Magnetic resonance imaging (MRI) has been extensively explored with deep neural networks to estimate brain age. The discrepancy between the predicted age and chronological age (real age) can be instrumental in identifying brain-related issues and assessing overall brain health. In this study, we have developed a brain age estimation framework utilizing a ResNet-50 deep neural network and a universum learning-based kernel random vector functional link (UKRVFL) network based on MRI images. A novel formulation of universum-KRVFL is introduced for regression tasks that capitalizes on prior knowledge through supplementary data samples. The universum data samples originate from the same domain as training samples but have different distributions. The proposed work efficacy is substantiated by conducting experiments on publicly available datasets. The model performance is quantified through metrics such as the mean absolute error (MAE), root mean square error (RMSE), and the coefficient of determination ( R 2 ), where lower MAE and RMSE values and a higher R 2 indicate greater accuracy in age prediction. The proposed age prediction model achieved an MAE of 2.68 years and 3.53 years of RMSE on healthy control (HC) test subjects. To further assess the significance of the brain age gap (BAG) as a biomarker for brain health, studies are conducted on mild cognitive impairment (MCI), PD, and AD groups. For MCI, PD, and AD groups, age estimation model yielded an RMSE of 4.13, 4.86, and 6.60 years, respectively. The experimental results demonstrate that the brain age gap in brain function is notably wider within AD group, indicating an acceleration of brain aging among those with neurodegeneration. [ABSTRACT FROM AUTHOR]

Patents contain detailed information of researched and developed technologies. We analyzed patent documents to understand the technology in a given domain. For the patent data analysis, we extracted the keywords from the patent documents using text mining techniques. Next, we built a patent document–keyword matrix using the patent keywords and analyzed the matrix data using statistical methods. Each element of the matrix represents the frequency of a keyword that occurs in a patent document. In general, most of the elements were zero because the keyword becomes a column of the matrix even if it occurs in only one document. Due to this zero-inflated problem, we experienced difficulty in analyzing patent keywords using existing statistical methods such as linear regression analysis. The purpose of this paper is to build a statistical model to solve the zero-inflated problem. In this paper, we propose a regression model based on quantile cumulative distribution function to solve this problem that occurs in patent keyword analysis. We perform experiments to show the performance of our proposed method using patent documents related to blockchain technology. We compare regression modeling based on a quantile cumulative distribution function with convenient models such as linear regression modeling. We expect that this paper will contribute to overcoming the zero-inflated problem in patent keyword analysis performed in various technology fields. [ABSTRACT FROM AUTHOR]

In recent years, there has been a focused effort to reduce the harmful effects of synthetic and mineral-based lubricants by emphasizing the use of biodegradable-based lubricants. These lubricants play a crucial role in minimizing friction, ensuring smooth operation of machines, and reducing the likelihood of frequent failures. With petroleum-based reserves depleting worldwide, prices are rising, and environmental damage is increasing. However, biolubricants derived from nonedible vegetable oils offer environmental benefits as they are nontoxic, emit minimal greenhouse gases, and are biodegradable. In this study, biolubricants are synthesized from jatropha and jojoba oil using sulphuric acid (H2SO4) and hydrochloric acid (HCl) as catalysts through the transesterification and epoxidation processes. The optimization of influencing parameters is achieved using Taguchi's orthogonal array, a statistical methodology. By employing design of experiments (DOE), the number of experimental trials is minimized while providing comprehensive details on the impact of control factors such as molar ratio, catalyst concentrations, and temperature. The results obtained from DOE reveal that the best optimized yield for jatropha biolubricant with H2SO4 and HCl catalysts is achieved with a molar ratio of 0.5:1.5, a temperature of 70 °C, and a catalyst concentration of 1.2 ml. The experimental yield for jatropha biolubricant with H2SO4 and HCl catalysts was measured at 226 ml and 238 ml, respectively, while the model predicted yield was 221 ml and 231 ml, respectively. The experimental yield for jojoba biolubricant with H2SO4 and HCl catalysts was recorded at 232 ml and 248 ml respectively, whereas the model predicted yield was 226 ml and 245 ml, respectively. Based on the analysis of variance (ANOVA) results, it is evident that among the three control factors, the molar ratio significantly influences the yield of both jatropha and jojoba biolubricants, as indicated by a p-value of less than 5%. The percentage contribution of the molar ratio in jatropha biolubricant with H2SO4 and HCl catalysts is found to be 98.99% and 97.2%, respectively. Furthermore, the R2 value, which exceeds 90%, signifies a strong relationship between the independent and dependent variables. The deviation between the experimental and regression-predicted equations for the yield remains within 2.5% for all combinations of jatropha and jojoba biolubricants. In conclusion, the study successfully prepared biolubricants from jatropha and jojoba-based non-edible vegetable oils and determined the optimal conditions for their production. [ABSTRACT FROM AUTHOR]

The objective of this research is to provide a comprehensive analysis of diamond price prediction by evaluating a wide array of 23 machine learning models (ML), including both regression and classification techniques. This study aims to fill a gap in existing literature by applying hyperparameter tuning optimization across various models to enhance prediction accuracy, estimated values and Time execution efficiency, setting a new benchmark in the field. This approach involved a systematic assessment of multiple ML models on their base and tuned performance concerning accuracy, execution time, and predictive value alignment (under, accurate, over). The study utilized advanced hyperparameter tuning techniques to optimize each model's performance, offering a comparative analysis that highlights the effectiveness of different models in predicting diamond prices. This research makes a distinct contribution through its extensive benchmarking of numerous ML models in the context of diamond price prediction, which is unprecedented in the literature. By applying hyperparameter tuning extensively to enhance model performance, its originality is derived from its comprehensive application of hyperparameter tuning to improve model performance by essentially tuning the model, this paper provides a novel contribution to the growing area of predictive analytics. By benchmarking an unprecedented amount of ML models for diamond price prediction and employing hyperparameter tuning, this paper moves the state of the art by noting the remarkable scope for accuracy improvements in tailored ML applications and demonstrates the extreme importance of model selection and optimization. The findings encompass that CatBoost Regressor, XGBoost Regressor still, kept high accuracy scores after tuning process and Random Forest Regressor accelerated much after tuning. Lastly, CatBoost Classifier, LightGBM Classifier existent achieving accuracies and efficiencies on the problem of diamond price classification tasks. Given its holistic nature, this study acknowledges the potential of overfitting in highly tuned models and their reliance on the specific dataset used for training. Future research might explore the generalisability of these techniques to other datasets and further investigate the trade-offs between model complexity and interpretability. The practical implications of this research are significant for stakeholders in the diamond industry such as retailers, appraisers, and investors. By identifying the most effective models for price prediction, we offer actionable insights that can improve decision-making processes, optimize inventory management, and enhance pricing policies. [ABSTRACT FROM AUTHOR]

In many different fields and industries, count data are publicly accessible. Control charts are used in quality studies to track count procedures. These control charts, however, only have a limited impact on zero-inflated data that contains extra zeros. The Zero-inflated power series (ZIPS) models, particularly its crucial sub-models, the Zero-inflated Poisson (ZIP), the Zero-inflated Negative binomial (ZINB), and the Zero-inflated Logarithmic (ZIL) models, are crucial approaches to handle the count data, and some control charts based on them have been proposed. However, there are situations when inflation can happen at one or more points other than zero (for instance, at one) or at more than one point (for instance, zero, one, and two). In these situations, the family of zero to k inflated power series (ZKIPS) models must be used in the control. In this work, we use a weighted score test statistic to examine upper-sided Shewhart, exponentially weighted moving average, and exponentially weighted moving average control charts. We only conducted numerical experiments on the zero to k Poisson model, which is one of the zero to k power series models, as an example. In ZKIPS models, the exponentially weighted moving average control chart can identify positive changes in the basis distribution's characteristics. By adding random effects, this method, in particular, enables boosting the capability of detecting unnatural heterogeneity variability. For detecting small to moderate shifts, the proposed strategy is more effective than the current Shewhart chart, according to simulation findings obtained using the Monte Carlo methodology. To show the charts' usefulness, they are also applied to a real example. [ABSTRACT FROM AUTHOR]

We have attempted to estimate blood glucose levels based on facial images measured in the near-infrared band, which is highly biopermeable, to establish a remote minimally invasive blood glucose measurement method. We measured facial images in the near-infrared wavelength range of 760–1650 nm, and constructed a general model for blood glucose level estimation by linear regression using the weights of spatial features of the measured facial images as explanatory variables. The results showed that the accuracy values of blood glucose estimation in the generalization performance evaluation were 43.02 mg/dL for NIR-I (760–1100 nm) and 43.61 mg/dL for NIR-II (1050–1650 nm) in the RMSE of the general model. Since biological information is nonlinear, it is necessary to explore suitable modeling methods for blood glucose estimation, including not only linear regression but also nonlinear regression. The purpose of this study is to explore suitable regression methods among linear and nonlinear regression methods to construct a blood glucose estimation model based on facial images with wavelengths from 760 to 1650 nm. The results showed that model using Random Forest had the best estimation accuracy with an RMSE of 36.02 mg/dL in NIR-I and the MR model had the best estimation accuracy with RMSE of 36.70 mg/dL in NIR-II under the current number of subjects and measurement data points. The independent components selected for the model have spatial features considered to be simply individual differences that are not related to blood glucose variation. [ABSTRACT FROM AUTHOR]

In the past few years, machine learning (ML) has become a widely used computational tool in manufacturing. The present study conducted a comparative analysis of five machine learning (ML) algorithms to predict the bead geometry in wire-arc additive manufacturing (WAAM). A robotic-controlled gas metal arc welding (GMAW) based WAAM setup was employed to deposit the beads. For the purpose of training and testing the ML models, a dataset consisting of four independent variables (travel speed, torch angle, wire feed rate, and contact tube to workpiece distance) and two dependent variables (bead height and width) was collected. The study employed five ML models including decision tree, random forest, XGBoost algorithm, linear regression, and artificial neural networks (ANN). The performance of ML models was evaluated using statistical metrics such as index of merit (IM), mean square error (MSE), mean absolute error (MAE) and coefficient of determination (R2 value). A comparison of ML algorithms revealed that ANN (R2 value = 0.6505) performed better than the XGBoost (R2 value = 0.5916) in predicting the bead height. The lowest value of MSE and MAE equal to 0.0611 and 0.1893 respectively was achieved for ANN in predicting the bead height. For the prediction of bead width, linear regression (R2 value = 0.8486) performed better than ANN (R2 value = 0.8316). The lowest value of MSE and MAE equal to 0.1730 and 0.3157 respectively was achieved for linear regression in predicting the bead width. Linear regression is most suitable for datasets having strong collinearity, however, ANN, Random Forest and XGBoost performed very well in prediction when the dataset shows high non-linear relations. The comparison of ML algorithms revealed that computational work using machine learning gives high-performance results with minimal use for resources and time. Hence, machine learning can be the most effective computational tool in modern manufacturing industries. [ABSTRACT FROM AUTHOR]

Purpose: To evaluate the 15 year incidence and progression of Diabetic Retinopathy (DR) and identify risk factors among Indian population.Methods: From a cross-sectional study of 1425 subjects, 911 participants took part in the 4-year follow-up. Out of these 911 participants, 140 returned for the 15-year follow-up, with baseline examinations conducted between 2003 and 2006, and subsequent follow-ups occurring from 2007 to 2011 and the current 15-year follow-up from 2018 to 2021. Of the 140 participants, 112 were eligible for analysis after excluding individuals with ungradable fundus photographs.Results: The 15-year incidence of any diabetic retinopathy (DR) was 5%, with mild NPDR and moderate NPDR at 1.57% and 2.7%, respectively. Proliferative DR was observed in 0.71% of cases, while diabetic macular edema (DME) and sight-threatening diabetic retinopathy (STDR) rates were 0.48% and 1.10%, respectively. Age-standardized rates revealed a significant association with increasing age and incident any DR and STDR. DR progression over 15 years included 7.5% one-step and 1.75% two-step progressions, while regression was limited to 1.75% one-step regression. Multiple logistic regression analyses revealed that baseline duration of diabetes, systolic blood pressure, HbA1c levels, and the presence of anemia influenced the incidence of any DR, DME, and STDR. Smoking and higher HbA1c were identified as risk factors for one-step progression of DR.Conclusion: This study provides crucial insights into the long-term incidence, progression, and regression of DR among individuals with Type 2 diabetes in India. [ABSTRACT FROM AUTHOR]

The aim of the study was to verify whether selected somatic parameters and components of body composition were significant predictors of cardiorespiratory fitness (CRF) among a potentially healthy Polish population of adolescents aged 11–14 years. The cross-sectional study was conducted on a group of 375 subjects (164 girls, and 211 boys). A 20 m shuttle run test (20 m SRT) was used to assess CRF. The total number of rounds was taken into account. Basic somatic parameters were measured: body mass (BM), body height (BH), waist circumference (WC), hip circumference (HC), body mass index (BMI), waist-to-hip ratio (WHR) and waist-to-height ratio (WHtR), and body composition components: body fat percentage (FM%), fat mass (FM kg), total body water (TBW), fat-free mass (FFM). Statistical analyses included basic statistical measures (mean and standard deviation) and Spearman rank correlation coefficient. Multiple linear regression analysis was performed to detect significant predictors of CRF. In each proposed model, the dependent variable was the number of rounds, and the independent variables were selected somatic parameters and components of body composition. More than half (65%) of the subjects had an average or lower level of CRF, and 35% of the population presented a good or above good level of CRF. The study showed a statistically significant negative correlation between BM, FM%, FM kg, HC, WC, BMI, WHR, WHtR and the number of laps in the total sample. The strongest correlation in the group of girls was noted for age (r = 0.34) and in the group of boys for FM% (r = - 0.52 ). Each regression model presented proved to be statistically significant, and the significant predictors of CRF in the group of girls were age ( R 2 = 16%) and FM% ( R 2 = 6%). In the group of boys, the significant predictors of CRF were WHtR ( R 2 = 8%) and age ( R 2 = 2%). Estimating body fat distribution is useful in assessing cardiorespiratory fitness, and this in turn indicates its usefulness in preventive screening of school-aged adolescents. [ABSTRACT FROM AUTHOR]