A Temperature-Based Statistical Model for Real-Time Thermal Deformation Prediction in End-Milling of Complex Workpiece

ASTRO-H is a next-generation X-ray observatory satellite planned for launch in fiscal 2015. Because ASTRO-H is a large structure and requires precise and accurate pointing, an understanding of on-orbit thermal deformation behavior is particularly important to the success of the ASTRO-H mission. In the conventional technique for predicting thermal deformation, testing is performed in a thermal vacuum chamber so asto simulate the temperature distribution on orbit. However, when using this conventional technique it is difficult to observe the detailed characteristics of the tested structures due to the limits of the applied temperature distribution cases. To overcome this problem, we propose a new evaluation technique for ASTRO-H. In our technique, on-orbit prediction of thermal deformation is performed thorough numerical simulation using a finite element (FE) model. The validation and correlation of the FE model are carried out by comparing results with thermal deformation test data collected in advance. This technique is especially suitable for satellites that are exposed to various thermal conditions on orbit. Using this technique, we confirmed all of the requirements for on-orbit pointing accuracy and precision in ASTRO-H.

In this study, thermal deformation in non-planar, large-scale additive manufacturing (LSAM) was experimentally and numerically investigated. A Bézier-based non-planar build surface was fabricated by CNC machining, and a single layer of ABS was deposited using a hybrid LSAM system. Toolpaths with raster angles of 0° and 45° were generated for surface-conformal printing. Infrared thermography was employed to monitor the thermal history during deposition. A three-dimensional finite element model was developed to simulate transient heat transfer and thermally induced deformation. Experimental deformation was quantified by 3D scanning and compared with simulation results. The results show that the slope geometry strongly influences deformation direction: negative slopes promote contraction, whereas positive slopes lead to upward deflection. Maintaining the material temperature above the glass transition temperature significantly reduces skew deformation. The finite element method predictions demonstrate strong agreement with experimental measurements, with normalized root mean square errors (NRMSEs) of approximately 11% for thermal deformation and 10% for temperature history. The proposed framework enables prediction and mitigation of thermal warping in non-planar polymer additive manufacturing.

Over-fitting is a ubiquitous problem in machine learning, and a variety of techniques to avoid over-fitting the training sample have proven highly effective, including early stopping, regularization, and ensemble methods. However, while over-fitting in training is widely appreciated and its avoidance now a standard element of best practice, over-fitting can also occur in model selection. This form of over-fitting can significantly degrade generalization performance, but has thus far received little attention. For example the kernel and regularization parameters of a support vector machine are often tuned by optimizing a cross-validation based model selection criterion. However the cross-validation estimate of generalization performance will inevitably have a finite variance, such that its minimizer depends on the particular sample on which it is evaluated, and this will generally differ from the minimizer of the true generalization error. Therefore if the cross-validation error is aggressively minimized, generalization performance may be substantially degraded. In general, the smaller the amount of data available, the higher the variance of the model selection criterion, and hence the more likely over-fitting in model selection will be a significant problem. Similarly, the more hyper-parameters to be tuned in model selection, the more easily the variance of the model selection criterion can be exploited, which again increases the likelihood of over-fitting in model selection. Over-fitting in model selection is empirically demonstrated to pose a substantial pitfall in the application of kernel learning methods and Gaussian process classifiers. Furthermore, evaluation of machine learning methods can easily be significantly biased unless the evaluation protocol properly accounts for this type of over-fitting. Fortunately the common solutions to avoiding over-fitting in training also appear to be effective in avoiding over-fitting in model selection. Three examples are presented based on regularization of the model selection criterion, early stopping in model selection and minimizing the number of hyper-parameters to be tuned during model selection.

This study aimed to determine the stability (in terms of covariate selection) of a population pharmacokinetic model and evaluate its performance in the absence of a test data set. Data from 88 full-term infants, 11 of whom were human immunodeficiency virus (HIV)-seropositive, taking an antiinfective agent were analyzed using exploratory data analysis methods and the nonlinear mixed-effects modeling (NONMEM) program to obtain the final population pharmacokinetic model. The stability of the population pharmacokinetic model was tested using the nonparametric bootstrap approach in four steps: 1) with the base pharmacokinetic model, 100 bootstrap replicates of the original data were generated by sampling with replacement; 2) ascertainment that each bootstrap data replicate was described by the basic structural model using the NONMEM objective function; 3) generalized additive modeling (GAM) applied to empiric Bayesian estimates for covariate selection at alpha = 0.05 and a frequency (f) cutoff value of 0.50; and 4) NONMEM population model building using covariates selected in the third step with alpha = 0.005. Performance of the population pharmacokinetic model was evaluated using 200 additional bootstrap replicates of the data by fitting the model obtained in step 4 to them. Parameters obtained were compared with those obtained in the model stability step, and improved prediction error, a measure of predictive accuracy as an index of internal validation, was computed. The reciprocal of serum creatinine (RSC; f = 0.73) and HIV (f = 0.70) were selected by GAM as predictors of clearance (Cl). The population pharmacokinetic model obtained without the determination of model stability included RSC as a predictor of Cl, but the final model from the model stability step included both HIV and RSC as predictors of Cl. Final population pharmacokinetic parameters were obtained with this model fitted to the original data; however, the 95% confidence interval on the HIV status regression coefficient included zero, indicating no significance. The mean parameter estimates obtained with the additional 200 bootstrap replicates of data were within 15% of those obtained with the final model at the regression stability step. Bootstrap resampling procedure is useful for evaluating the stability and performance of a population model by repeatedly fitting it to the bootstrap samples when there is no test data set.

The advancement of urbanization has led to stricter requirements for the prediction and control of excavation-induced deformations. To achieve this goal, this study proposes a novel method that integrates a spatiotemporal graph attention network (ST-GAT) with a proportional–integral–derivative (PID) controller to proactively control wall deflections caused by excavation. The ST-GAT model improves wall deflection prediction by capturing spatial relationships between monitoring points near steel struts and dynamically assigning weights based on their importance. The interpretability of the model is greatly improved by generating a feature attribution map across various input features and visualizing the weight distribution between nodes in the GAT network. A proactive control method of wall deflections is proposed by replacing current monitoring values in the PID control system with predicted values for multiple steel struts using the ST-GAT model. Compared to the standard PID method, this approach can control wall deflections before significant deformations occur. A real excavation project equipped with a servo support system is used to validate the effectiveness of the proposed method. The results show that the ST-GAT model significantly outperforms other models, and its performance improves when utilizing spatial relationships from more monitoring points. With a reasonable combination of PID hyperparameters, the proposed ST-GAT-based PID controller can control wall deflections close to a target value.

Background: Primary biliary cholangitis (PBC) is a rare autoimmune liver disease with few effective treatments and a poor prognosis, and its incidence is on the rise. There is an urgent need for more targeted treatment strategies to accurately identify high-risk patients. The use of stochastic survival forest models in machine learning is an innovative approach to constructing a prognostic model for PBC that can improve the prognosis by identifying high-risk patients for targeted treatment. Method: Based on the inclusion and exclusion criteria, the clinical data and follow-up data of patients diagnosed with PBC-associated cirrhosis between January 2011 and December 2021 at Taizhou Hospital of Zhejiang Province were retrospectively collected and analyzed. Data analyses and random survival forest model construction were based on the R language. Result: Through a Cox univariate regression analysis of 90 included samples and 46 variables, 17 variables with p-values <0.1 were selected for initial model construction. The out-of-bag (OOB) performance error was 0.2094, and K-fold cross-validation yielded an internal validation C-index of 0.8182. Through model selection, cholinesterase, bile acid, the white blood cell count, total bilirubin, and albumin were chosen for the final predictive model, with a final OOB performance error of 0.2002 and C-index of 0.7805. Using the final model, patients were stratified into high- and low-risk groups, which showed significant differences with a P value <0.0001. The area under the curve was used to evaluate the predictive ability for patients in the first, third, and fifth years, with respective results of 0.9595, 0.8898, and 0.9088. Conclusion: The present study constructed a prognostic model for PBC-associated cirrhosis patients using a random survival forest model, which accurately stratified patients into low- and high-risk groups. Treatment strategies can thus be more targeted, leading to improved outcomes for high-risk patients.

Introduction: Monitoring of AUC24 was updated recommendation in the guideline for the therapeutic drug monitoring (TDM) of vancomycin in Chinese pharmacological society published in 2020. Vancomycin pharmacokinetic profiles are diverse and unique in critically ill patients because of the drastic variability of the patients’ physiological parameters, while the study for population pharmacokinetic (PPK) models in Chinese critically ill patients has been rarely reported. The objectives of this study were to construct a PPK model to describe the pharmacokinetic characteristics of vancomycin in critically ill patients and to individualize vancomycin dosing by model-informed Bayesian estimation for maintenance of AUC24 target at 400–650 mg h/L recommended by the 2020 guideline. Methods: Vancomycin with different dosing was administered intravenously over 1 h for critically ill patients, TDM was started at 48 h or 72 h since initiation of vancomycin therapy for patients. Blood samples were collected from patients for trough concentrations or Cmax. Vancomycin concentrations were determined by high-performance liquid chromatography method with ultraviolet detection. PPK model was performed using the nonlinear mixed-effect model (NONMEM®). Individual PK parameters for critically ill patients treated with vancomycin were estimated using a post hoc empirical Bayesian method based on the final PPK model. AUC24 was calculated as the total daily dose divided by the clearance (L/h). Results: The PPK of vancomycin was determined by a one-compartment model with creatinine clearance as fixed effects. The PK estimates in the final model generally agreed with the median estimates and were contained within the 95% CI generated from the bootstrap results, indicating good precision and stability in the final model. The visual predictive check plots showed the adequate predictive performance of the final PK model and supported a good model fit. The model-informed Bayesian estimation was used to predict the AUC24 of critically ill patient by the acquired TDM results, and the dosing adjustment by maintenance of AUC24 at 400–650 mg h/L had made a great therapeutic effect for the case. Conclusion: This study established a PPK model of vancomycin in Chinese critically ill patients, and individualized dosing of vancomycin by model-informed Bayesian estimation to maintain an AUC24 target at 400–650 mg h/L has been successfully applied in clinic. This result supports the continued use of model-informed Bayesian estimation to vancomycin treatment in critically ill patients.

In the design of folding table, we should consider three evaluation indexes: good product stability, easy processing and least material. The stability of the product is characterized by the angle of tilt of the legs and the distance between the legs of the table through force analysis; the convenience of processing is characterized by minimizing the maximum slot length. Therefore, based on these three indicators, a multi-objective programming model is established. In solving the model, the weight with good stability, convenient processing and least material consumption is obtained by AHP, and the multi-objective programming is transformed into single-objective programming. For the given height of the folding table and the diameter of the circular table, the optimum design and machining parameters such as the position of the steel bar and the size of the plate are obtained according to the model.

Machine learning and artificial intelligence techniques have been applied to construct online portfolio selection strategies recently. A popular and state-of-the-art family of strategies is to explore the reversion phenomenon through online learning algorithms and statistical prediction models. Despite gaining promising results on some benchmark datasets, these strategies often adopt a single model based on a selection criterion (e.g., breakdown point) for predicting future price. However, such model selection is often unstable and may cause unnecessarily high variability in the final estimation, leading to poor prediction performance in real datasets and thus non-optimal portfolios. To overcome the drawbacks, in this article, we propose to exploit the reversion phenomenon by using combination forecasting estimators and design a novel online portfolio selection strategy, named Combination Forecasting Reversion (CFR), which outputs optimal portfolios based on the improved reversion estimator. We further present two efficient CFR implementations based on online Newton step (ONS) and online gradient descent (OGD) algorithms, respectively, and theoretically analyze their regret bounds, which guarantee that the online CFR model performs as well as the best CFR model in hindsight. We evaluate the proposed algorithms on various real markets with extensive experiments. Empirical results show that CFR can effectively overcome the drawbacks of existing reversion strategies and achieve the state-of-the-art performance.

The deformation of concrete dams directly reflects their structural health and operational state, serving as a critical foundation for safety assessment and early risk warning. Accurately predicting dam deformation patterns is thus essential for ensuring long‐term structural safety and enabling scientific operation management. However, existing models remain limited in addressing high‐frequency noise in monitoring data, performing dynamic feature selection, and modeling complex spatiotemporal dependencies, which collectively constrain prediction accuracy. To overcome these challenges, this study proposes a dam deformation prediction model that integrates variational mode decomposition with wavelet thresholding (VMD–WT), a partial autocorrelation function (PACF)–based dynamic feature selection approach, and the ScaleGraph Block‐Mamba‐like linear attention (SGB–MLLA) –Transformer. The proposed model performs multiscale signal decomposition to suppress noise and extract dominant deformation trends, while dynamically selecting key influencing factors and incorporating spatial dependency modeling and lightweight attention mechanisms to enhance the representation of long sequence and multifactor coupled deformation features. To validate the model’s effectiveness, deformation data from monitoring points of a concrete dam in Jiangxi Province, China, were used for evaluation. Experimental results demonstrate that the proposed model achieves superior prediction performance across multiple monitoring points, achieving near‐perfect accuracy ( R 2 = 0.9993) with submillimeter error margins at GLD4, significantly outperforming existing models. These findings confirm that integrating frequency‐domain decomposition with adaptive feature selection and employing linear attention for efficient long sequence modeling can substantially improve deformation prediction accuracy. This research provides a novel methodological framework for dam health diagnosis and safety management, offering both theoretical and practical value for the development of intelligent dam monitoring systems.

Increasingly, deformation prediction has become an essential research topic in sluice safety control, which requires significant attention. However, there is still a lack of practical and efficient prediction modeling for sluice deformation. In order to address the limitations in mining the deep features of long-time data series of the traditional statistical model, in this paper, an improved long short-term memory (LSTM) model and weighted Markov model are introduced to predict sluice deformation. In the method, the seagull optimization algorithm (SOA) is utilized to optimize the hyper-parameters of the neural network structure in LSTM primarily to improve the model. Subsequently, the relevant error sequences of the fitting results of SOA-LSTM model are classified and the Markovity of the state sequence is examined. Then, the autocorrelation coefficients and weights of each order are calculated and the weighted and maximum probability values are applied to predict the future random state of the sluice deformation. Afterwards, the prediction model of sluice deformation on the SOA-LSTM-weighted Markov model is proposed. Ultimately, the presented model is used to predict the settlement characteristics of an actual sluice project in China. The analysis results demonstrate that the proposed model possesses the highest values of R2 and the smallest values of RMSE and absolute relative errors for the monitoring data of four monitoring points. Consequently, it concluded that the proposed method shows better prediction ability and accuracy than the SOA-LSTM model and the stepwise regression model.

In causal inference there are often multiple reasonable estimators for a given target quantity. For example, one may reasonably use inverse probability weighting, an instrumental variables approach, or construct an estimate based on proxy outcomes if the actual outcome is difficult to measure. Ideally, the practitioner decides on an estimator before looking at the data. However, this might be challenging in practice since a priori it might not be clear to a practitioner how to choose the method. If the final model is chosen after peeking at the data, naive inferential procedures may fail. This raises the need for a model selection tool, with rigorous asymptotic guarantees. Since there is usually no loss function available in causal inference, standard model selection techniques do not apply. We propose a model selection procedure that estimates the squared ℓ2-deviation of a finite-dimensional estimator from its target. The procedure relies on knowing an asymptotically unbiased (potentially highly variable) estimate of the parameter of interest. The resulting estimator is discontinuous and does not have a Gaussian limit distribution. Thus, standard asymptotic expansions do not apply. We derive asymptotically valid confidence intervals for low-dimensional settings that take into account the model selection step. The performance of the approach for estimation and inference for average treatment effects is evaluated on simulated data sets in low-dimensional settings, including experimental data, instrumental variables settings and observational data with selection on observables.

Mechanics of bone graft and implant choices for spino-pelvic reconstruction following combined hemipelvectomy, sacrectomy and L5 vertebrectomy.

Welders are frequently exposed to Manganese (Mn), which may increase the risk of neurological impairment. Historical exposure estimates for welding-exposed workers are needed for epidemiological studies evaluating the relationship between welding and neurological or other health outcomes. The objective of this study was to develop and validate a multivariate model to estimate quantitative levels of welding fume exposures based on welding particulate mass and Mn concentrations reported in the published literature. Articles that described welding particulate and Mn exposures during field welding activities were identified through a comprehensive literature search. Summary measures of exposure and related determinants such as year of sampling, welding process performed, type of ventilation used, degree of enclosure, base metal, and location of sampling filter were extracted from each article. The natural log of the reported arithmetic mean exposure level was used as the dependent variable in model building, while the independent variables included the exposure determinants. Cross-validation was performed to aid in model selection and to evaluate the generalizability of the models. A total of 33 particulate and 27 Mn means were included in the regression analysis. The final model explained 76% of the variability in the mean exposures and included welding process and degree of enclosure as predictors. There was very little change in the explained variability and root mean squared error between the final model and its cross-validation model indicating the final model is robust given the available data. This model may be improved with more detailed exposure determinants; however, the relatively large amount of variance explained by the final model along with the positive generalizability results of the cross-validation increases the confidence that the estimates derived from this model can be used for estimating welder exposures in absence of individual measurement data.

An empirical approach to model selection through validation for censored survival data