Search

This document is a correction notice for an article titled "Cost and statistical efficiency of posture assessment by inclinometry and observation, exemplified by paper mill work." The correction is related to an error in Figure 2, specifically the efficiency ordinate scale in panel B. The corrected figure is provided in the document. The article discusses the sampling efficiency and costs associated with different strategies for posture assessment using inclinometry and observation in the context of paper mill work. The authors of the article are Svend Erik Mathiassen, Amanda Waleh Åström, Annika Strömberg, and Marina Heiden. [Extracted from the article]

The vibration of a transformer core induced by magnetostriction is closely related to the environment and safety of devices. The DC bias excitation leads to the over saturation of the transformer core and causes more severe vibration. In this paper, a magnetic-mechanical strong coupling model based on a fixed-point time-periodic 2D finite element method with excitation voltage and DC bias current as the input is established in which the vector magnetic potential A, winding current I, and displacement in the core u are simultaneously taken as solution variables, and the fixed-point reluctivity is introduced to deal with the nonlinear problem of the core. In this model, the magnetic field and mechanical field are associated with the piezomagnetic equations and variational principle, and the circuit and magnetic fields are connected through the law of electromagnetic induction. The equivalent thickness of the core model and the determination of the initial value are discussed. A test transformer with a laminated core is made for experimental measurement. The accuracy and efficiency of the model proposed are verified by comparing the calculated results with the experimental results, and the influence of DC bias components on the magnetic characteristics and vibration of the transformer core is analyzed. [ABSTRACT FROM AUTHOR]

A new methodology for solving the fuzzy multiobjective optimization problems is proposed in this paper by considering the fusion of cooperative game theory and genetic algorithm. The original fuzzy multiobjective optimization problem needs to be transformed into a scalar optimization problem, which is a conventional optimization problem. Usually, the assignments of suitable coefficients to the corresponding scalar optimization problem are subjectively determined by the decision makers. However, these assignments may cause some biases by their subjectivity. Therefore, this paper proposes a mechanical procedure to avoid this subjective biases. We are going to formulate a cooperative game using the α -level functions of the multiple fuzzy objective functions. Under this setting, the suitable coefficients can be determined mechanically by involving the core values of the cooperative game, which is formulated using the multiple fuzzy objective functions. We shall prove that the optimal solutions of the transformed scalar optimization problem are indeed the nondominated solutions of fuzzy multiobjective optimization problem. Since the core-nondominated solutions will depend on the coefficients that are determined by the core values of cooperative game, there will be a lot of core-nondominated solutions that will also depend on the corresponding coefficients. In order to obtain the best core-nondominated solution, we shall invoke the genetic algorithms by evolving the coefficients. [ABSTRACT FROM AUTHOR]

The linear exponential distribution is a generalization of the exponential and Rayleigh distributions. This distribution is one of the best models to fit data with increasing failure rate (IFR). But it does not provide a reasonable fit for modeling data with decreasing failure rate (DFR) and bathtub shaped failure rate (BTFR). To overcome this drawback, we propose a new record-based transmuted generalized linear exponential (RTGLE) distribution by using the technique of Balakrishnan and He. The family of RTGLE distributions is more flexible to fit the data sets with IFR, DFR, and BTFR, and also generalizes several well-known models as well as some new record-based transmuted models. This paper aims to study the statistical properties of RTGLE distribution, like, the shape of the probability density function and hazard function, quantile function and its applications, moments and its generating function, order and record statistics, Rényi entropy. The maximum likelihood estimators, least squares and weighted least squares estimators, Anderson-Darling estimators, Cramér-von Mises estimators of the unknown parameters are constructed and their biases and mean squared errors are reported via Monte Carlo simulation study. Finally, the real data sets illustrate the goodness of fit and applicability of the proposed distribution; hence, suitable recommendations are forwarded. [ABSTRACT FROM AUTHOR]

Multiple measures, such as WEAT or MAC, attempt to quantify the magnitude of bias present in word embeddings in terms of a single-number metric. However, such metrics and the related statistical significance calculations rely on treating pre-averaged data as individual data points and utilizing bootstrapping techniques with low sample sizes. We show that similar results can be easily obtained using such methods even if the data are generated by a null model lacking the intended bias. Consequently, we argue that this approach generates false confidence. To address this issue, we propose a Bayesian alternative: hierarchical Bayesian modeling, which enables a more uncertainty-sensitive inspection of bias in word embeddings at different levels of granularity. To showcase our method, we apply it to Religion, Gender, and Race word lists from the original research, together with our control neutral word lists. We deploy the method using Google, GloVe, and Reddit embeddings. Further, we utilize our approach to evaluate a debiasing technique applied to the Reddit word embedding. Our findings reveal a more complex landscape than suggested by the proponents of single-number metrics. The datasets and source code for the paper are publicly available. [ABSTRACT FROM AUTHOR]

Rapid developments in satellite remote-sensing technology have enabled the collection of geospatial data on a global scale, hence increasing the need for covariance functions that can capture spatial dependence on spherical domains. We propose a general method of constructing nonstationary, locally anisotropic covariance functions on the sphere based on covariance functions in R 3 . We also provide theorems that specify the conditions under which the resulting correlation function is isotropic or axially symmetric. For large datasets on the sphere commonly seen in modern applications, the Vecchia approximation is used to achieve higher scalability on statistical inference. The importance of flexible covariance structures is demonstrated numerically using simulated data and a precipitation dataset. Supplementary materials accompanying this paper appear online. [ABSTRACT FROM AUTHOR]

The recently proposed semantic-aligned matching detection transformer (SAM–DETR model) accelerates the convergence of the detection transformer (DETR) by mapping object queries into an identical embedding space as the encoder's output feature map. However, SAM–DETR model has the problem of low detection accuracy compared to other DETR variants. We observe that the lower detection accuracy of SAM–DETR model is caused by the insufficient number of sample points and the inaccurate localization of the sample points during re-sampling, which blurs the generated attention map. This paper proposes an object detector based on a feature correction and double sampling DETR (FCDS-DETR) to solve this problem. FCDS-DETR takes SAM–DETR model as a baseline and builds on it by adding a feature correction module and a double sampling mechanism to achieve further improvement in detection accuracy with a limited number of additional parameters without sacrificing convergence speed. Firstly, FCDS-DETR improves the sampling point localization accuracy by adding a feature correction module to model the inter-channel dependence of the feature maps to be sampled. Secondly, the number of sampled points is increased by the double sampling mechanism, and attention fusion is used to fuse the attention weight maps corresponding to the two sets of sampled points to improve the recognizability of the attention weight maps. The experimental results show that the average precision is improved by +0.7 on the COCO dataset compared with the SAM–DETR model, and the number of parameters is increased by only 10.34 % , which improves the detection performance of the model very well. [ABSTRACT FROM AUTHOR]

Fourier ptychographic microscopy, as a computational imaging method, can reconstruct high-resolution images but suffers optical aberration, which affects its imaging quality. For this reason, this paper proposes a network model for simulating the forward imaging process in the Tensorflow framework using samples and coherent transfer functions as the input. The proposed model improves the introduced Wirtinger flow algorithm, retains the central idea, simplifies the calculation process, and optimizes the update through back propagation. In addition, Zernike polynomials are used to accurately estimate aberration. The simulation and experimental results show that this method can effectively improve the accuracy of aberration correction, maintain good correction performance under complex scenes, and reduce the influence of optical aberration on imaging quality. [ABSTRACT FROM AUTHOR]

Background: Scattering photons can seriously contaminate cone‐beam CT (CBCT) image quality with severe artifacts and substantial degradation of CT value accuracy, which is a major concern limiting the widespread application of CBCT in the medical field. The scatter kernel deconvolution (SKD) method commonly used in clinic requires a Monte Carlo (MC) simulation to determine numerous quality‐related kernel parameters, and it cannot realize intelligent scatter kernel parameter optimization, causing limited accuracy of scatter estimation. Purpose: Aiming at improving the scatter estimation accuracy of the SKD algorithm, an intelligent scatter correction framework integrating the SKD with deep reinforcement learning (DRL) scheme is proposed. Methods: Our method firstly builds a scatter kernel model to iteratively convolve with raw projections, and then the deep Q‐network of the DRL scheme is introduced to intelligently interact with the scatter kernel to achieve a projection adaptive parameter optimization. The potential of the proposed framework is demonstrated on CBCT head and pelvis simulation data and experimental CBCT measurement data. Furthermore, we have implemented the U‐net based scatter estimation approach for comparison. Results: The simulation study demonstrates that the mean absolute percentage error (MAPE) of the proposed method is less than 9.72% and the peak signal‐to‐noise ratio (PSNR) is higher than 23.90 dB, while for the conventional SKD algorithm, the minimum MAPE is 17.92% and the maximum PSNR is 19.32 dB. In the measurement study, we adopt a hardware‐based beam stop array algorithm to obtain the scatter‐free projections as a comparison baseline, and our method can achieve superior performance with MAPE < 17.79% and PSNR > 16.34 dB. Conclusions: In this paper, we propose an intelligent scatter correction framework that integrates the physical scatter kernel model with DRL algorithm, which has the potential to improve the accuracy of the clinical scatter correction method to obtain better CBCT imaging quality. [ABSTRACT FROM AUTHOR]

Quantum vortices in Bose–Einstein condensates (BECs) are essential phenomena in condensed matter physics, and precisely locating their positions, especially the vortex core, is a precondition for studying their properties. With the rise of machine learning, there is a possibility to expedite the localization process and provide accurate predictions. However, traditional machine learning requires particular considerable amount of manual data annotation, leading to uncontrollable accuracy. In this paper, we utilize the U-Net method to detect vortex positions accurately at the pixel level and propose an Automatic Correction Labeling (ACL) approach to optimize the acquisition of data sets for vortex localization in BECs. This approach addresses inaccuracies in the labeled vortex positions and improves the accuracy of vortex localization, especially the vortex core positions, while enhancing the tolerance for human mislabeling. The main process involves Rough Labeling → Machine Learning → Probability Region Search → Data Relabeling → Machine Learning again. The objective of ACL is to secure more accurate labeled data for model retraining. Through vortex localization experiments conducted in a two-dimensional Bose-Einstein condensate, our results establish the following: 1. Even under conditions of biased and missing manual annotations, U-Net can still accurately locate vortex positions; 2. Vortices exhibit certain regularities, and training U-Net with a small number of samples yields excellent predictive consequences; 3. The machine learning vortex locator based on the ACL method effectively corrects errors in manually annotated data, significantly improving the model's performance metrics, thus enhancing the precision and metrics of vortex localization. This substantial advancement in the application of machine learning in vortex localization provides an effective way for vortex dynamics localization. Furthermore, this method of obtaining more accurate positions of approximate human labels through machine learning offers new insights for machine learning in other types of image recognition problems. [ABSTRACT FROM AUTHOR]

The June 2024 issue of Visual Computer includes six best papers from the Cyberworlds 2022 conference, five best papers from the second part of the special issue on Deep Learning for 3D Segmentation, and 34 regular papers. The selected papers from the Cyberworlds conference cover topics such as procedural modeling, context-aware personality estimation, immersive haptic simulation, video compression-cum-classification, semantics-aware addition and LoD of 3D window details, and military education in extended reality. The papers from the Deep Learning for 3D Segmentation section focus on point cloud semantic segmentation, structure-preserving image smoothing, multimodal feature fusion network, unsupervised contrastive learning, and 6D pose estimation of reflective texture-less objects. The regular papers cover a wide range of topics in computer graphics and visualization. The issue is edited by Nadia Magnenat Thalmann, the Editor-in-Chief of the Visual Computer. [Extracted from the article]

We correct a logic mistake in our paper "On statistical convergence and strong Cesàro convergence by moduli for double sequences" (León-Saavedra et al. in J. Inequal. Appl. 2022:62, 2022). [ABSTRACT FROM AUTHOR]

Background: Cone beam computed tomography (CBCT) plays an increasingly important role in image‐guided radiation therapy. However, the image quality of CBCT is severely degraded by excessive scatter contamination, especially in the abdominal region, hindering its further applications in radiation therapy. Purpose: To restore low‐quality CBCT images contaminated by scatter signals, a scatter correction algorithm combining the advantages of convolutional neural networks (CNN) and Swin Transformer is proposed. Methods: In this paper a scatter correction model for CBCT image, the Flip Swin Transformer U‐shape network (FSTUNet) model, is proposed. In this model, the advantages of CNN in texture detail and Swin Transformer in global correlation are used to accurately extract shallow and deep features, respectively. Instead of using the original Swin Transformer tandem structure, we build the Flip Swin Transformer Block to achieve a more powerful inter‐window association extraction. The validity and clinical relevance of the method is demonstrated through extensive experiments on a Monte Carlo (MC) simulation dataset and frequency split dataset generated by a validated method, respectively. Result: Experimental results on the MC simulated dataset show that the root mean square error of images corrected by the method is reduced from over 100 HU to about 7 HU. Both the structural similarity index measure (SSIM) and the universal quality index (UQI) are close to 1. Experimental results on the frequency split dataset demonstrate that the method not only corrects shading artifacts but also exhibits a high degree of structural consistency. In addition, comparison experiments show that FSTUNet outperforms UNet, Deep Residual Convolutional Neural Network (DRCNN), DSENet, Pix2pixGAN, and 3DUnet methods in both qualitative and quantitative metrics. Conclusions: Accurately capturing the features at different levels is greatly beneficial for reconstructing high‐quality scatter‐free images. The proposed FSTUNet method is an effective solution to CBCT scatter correction and has the potential to improve the accuracy of CBCT image‐guided radiation therapy. [ABSTRACT FROM AUTHOR]

In order to predict crop reference transpiration(ET0) scientifically and accurately, improve the prediction accuracy, and reduce the number of input variables, so as to reduce the construction cost of intelligent water-saving irrigation system, in this paper, deep learning and artificial neural network methods were used to establish ET0 intelligent prediction models, and local sensitivity analysis, fuzzy curve, and fuzzy surface methods were used to study the influence of each input variable in ET0 prediction on the prediction results. Based on the influence factor size, eight input combinations of different meteorological factors were constructed. The daily meteorological data of Rizhao weather station were used to train and test the prediction models with different methods and input variable combinations. The results of Penman formula were used as reference to evaluate the performance of the prediction models. The results showed that the R2 of the deep learning prediction model was 0.980, which was higher than that of the artificial neural network model which was 0.963 when the complete variables were used as the prediction input, and higher prediction accuracy was obtained. For ET0 prediction of default input variables, the performance of the deep learning prediction model was superior to that of the artificial neural network, and the R2 of the deep learning prediction model with the input parameters of average temperature and Sunshine duration still reached 0.935, indicating that the deep learning prediction model could still obtain good prediction results when there were only a few meteorological parameters. Through comprehensive analysis of R2, RMSE, RMSRE, MRE, MAPE, and other results, it could be seen that if the regional climate data of the study was limited, the deep learning model with input combinations of(n, T, RH, Tmin, Ws) and (n, T, RH, Ws) could be used to predict, compared with the results calculated by Penman formula, the MAPE values were 8.753 and 8.404, respectively, and the R2 values were both greater than 0.98, which could be used as a standard prediction model. [ABSTRACT FROM AUTHOR]

In this paper, the problem of the networked set-membership state estimation is discussed for a class of nonlinear discrete time-varying systems subject to cyber attacks and time delays. Two forms of malicious attacks (i.e. Denial-of-Service (DoS) attack and bias injection attack) are taken into account to describe the adversary's attempt to destroy/deteriorate the system performance via communication network. It is the aim of the investigated issue to propose a set-membership state estimator ensuring the required estimation performance despite the existence of both the external mixed attacks and internal time delays. By resorting to the feasibility of a series of matrix inequalities, sufficient conditions are provided for the solvability of the addressed state estimator design problem. Furthermore, an optimisation strategy is developed with the purpose of seeking the local optimal estimator parameters. At last, a numerical simulation example is presented to demonstrate the effectiveness of the proposed theoretical algorithm. [ABSTRACT FROM AUTHOR]

The bolt loosening detection method has been paid attention by engineering and academic scholars. The presented methods only focus on the identification of the bolt based on deep learning, and the problem of bolt localization and the influence of shadow on distortion correction is seldom studied. In this paper, a bolt self-localization method based on YOLOv4 deep learning algorithm is proposed. A bolt numbering rule is established and the deep learning is introduced to identify the number and locate the bolt. A new square bolt gasket is proposed and four corner points are used for distortion correction. In order to reduce the influence of shadows, a grayscale enhancement strategy is proposed to improve the correction stability. Finally, a laboratory flange joint is used to verify the proposed method. The results show that the bolt self-localization method is feasible and the new bolt gasket can effectively improve the stability of distortion correction and bolt loosening detection. [ABSTRACT FROM AUTHOR]

BeiDou Global Navigation Satellite System (BDS-3) broadcasts multifrequency signals that offer more choices of frequencies and more signal combinations for positioning. This paper analyzes the effect of timing group delay (TGD) and differential code bias (DCB) of BDS-3 on the corresponding triple-frequency positioning. The triple-frequency observation models of BDS-3 are summarized and the DCB correction models are derived for the four different frequency combinations of triple-frequency ionospheric-free (IF) combination (IF123), two dual-frequency IF combinations (IF1213) and triple-frequency uncombined (UC123) positioning modes. Standard point positioning (SPP) and precise point positioning (PPP) experiments were conducted using 30 days of observations from 25 multi-GNSS experiment (MGEX) stations. The results show that the TGD/DCB correction has a significant impact on the accuracy of SPP. The positioning accuracy using IF123 and IF1213 models improved by about 73~90% after TGD correction, in comparison to a 27~30% improvement achieved using the UC123 model. In addition, the correction effect of DCB is slightly better than TGD. The DCB correction significantly improves accuracy in the initial epoch of the PPP, which helps the convergence of the filtering and reduces the convergence time. The average convergence times of IF123, IF1213 and UC123 are 26.1, 26.9 and 38.3 min, respectively, which are reduced by 6.79, 2.54 and 8.59% with DCB correction. The pseudorange residuals are closer to zero-mean random noise after DCB correction. Furthermore, the DCB affects the evaluation of the inter-frequency bias (IFB), ionospheric delay and floating ambiguity parameters. However, the tropospheric delay is almost unaffected by DCB. [ABSTRACT FROM AUTHOR]

In this paper, we propose a total fractional-order variation model for multiplicative noise removal and contrast enhancement of real SAR images. Inspired by the high dynamic intensity range of SAR images, the full content of the SAR images is preserved by normalizing the original data in this model. Then, we propose a degradation model based on the nonlinear transformation to adjust the intensity of image pixel values. With MAP estimator, a corresponding fidelity term is introduced into the model, which is beneficial for contrast enhancement and bias correction in the denoising process. For the regularization term, a gray level indicator is used as a weighted matrix to make the model adaptive. We first apply the scalar auxiliary variable algorithm to solve the proposed model and prove the convergence of the algorithm. By virtue of the discrete Fourier transform (DFT), the model is solved by an iterative scheme in the frequency domain. Experimental results show that the proposed model can enhance the contrast of natural and SAR images while removing multiplicative noise. [ABSTRACT FROM AUTHOR]

Accurate medium- and long-term power load forecasting is of great significance for the scientific planning and safe operation of power systems. Monthly power load has multiscale time series correlation and seasonality. The existing models face the problems of insufficient feature extraction and a large volume of prediction models constructed according to seasons. Therefore, a hybrid feature pyramid CNN-LSTM model with seasonal inflection month correction for medium- and long-term power load forecasting is proposed. The model is constructed based on linear and nonlinear combination forecasting. With the aim to address the insufficient extraction of multiscale temporal correlation in load, a time series feature pyramid structure based on causal dilated convolution is proposed, and the accuracy of the model is improved by feature extraction and fusion of different scales. For the problem that the model volume of seasonal prediction is too large, a seasonal inflection monthly load correction strategy is proposed to construct a unified model to predict and correct the monthly load of the seasonal change inflection point, so as to improve the model's ability to deal with seasonality. The model proposed in this paper is verified on the actual power data in Shaoxing City. [ABSTRACT FROM AUTHOR]

Recently, there has been growing interest in using machine learning methods for causal inference due to their automatic and flexible ability to model the propensity score and the outcome model. However, almost all the machine learning methods for causal inference have been studied under the assumption of no unmeasured confounding and there is little work on handling omitted/unmeasured variable bias. This paper focuses on a machine learning method based on random forests known as Causal Forests and presents five simple modifications for tuning Causal Forests so that they are robust to cluster-level unmeasured confounding. Our simulation study finds that adjusting the default tuning procedure with the propensity score from fixed effects logistic regression or using variables that are centered to their cluster means produces estimates that are more robust to cluster-level unmeasured confounding. Also, when these parametric propensity score models are mis-specified, our modified machine learning methods remain robust to bias from cluster-level unmeasured confounders compared to existing parametric approaches based on propensity score weighting. We conclude by demonstrating our proposals in a real data study concerning the effect of taking an eighth-grade algebra course on math achievement scores from the Early Childhood Longitudinal Study. [ABSTRACT FROM AUTHOR]

The extreme value distribution was developed for modeling extreme-order statistics or extreme events. In this study, we discuss the distribution of the largest extreme. The main objective of this paper is to determine the best estimators of the unknown parameters of the extreme value distribution. Thus, both classical and Bayesian methods are used. The classical estimation methods under consideration are maximum likelihood estimators, moment's estimators, least squares estimators, and weighted least squares estimators, percentile estimators, the ordinary least squares estimators, best linear unbiased estimators, L-moments estimators, trimmed L-moments estimators, and Bain and Engelhardt estimators. We also propose new estimators for the unknown parameters. Bayesian estimators of the parameters are derived by using Lindley's approximation and Markov Chain Monte Carlo methods. The asymptotic confidence intervals are considered by using maximum likelihood estimators. The Bayesian credible intervals are also obtained by using Gibbs sampling. The performances of these estimation methods are compared with respect to their biases and mean square errors through a simulation study. The maximum daily flood discharge (annual) data sets of the Meriç River and Feather River are analyzed at the end of the study for a better understanding of the methods presented in this paper. [ABSTRACT FROM AUTHOR]

The statically indeterminate characteristics of arch dams highlight the temperature deformation effect, making accurate modelling of this effect a key issue in improving the performance of displacement monitoring models. In this paper, causal interpretation ability and prediction accuracy of five kinds of temperature deformation modelling factors, including seasonal harmonic function, segmented average previous air temperature, air temperature hysteresis correction factor, principal components and shape feature clustering-based principal components of measured dam temperatures, are compared. On this basis, a combination prediction model is established using the above five causal models as submodels. The combination process is conducted by three methods of dynamic mutual information coefficient, random forest and support vector machine. Research results of the Jinping-I arch dam show that the shape feature clustering-based temperature principal components can significantly improve the accuracy and adaptability of displacement monitoring models, in which the root mean square error decreases with an average rate of 52%. The combination prediction model can effectively take the advantages of different kinds of temperature deformation modelling factors into account. Compared with the hydraulic-seasonal-time model and the best submodel, prediction accuracy of the support vector machine-based combination model is improved with an average rate of 54% and 28%, respectively. [ABSTRACT FROM AUTHOR]

This paper deals with a new two-parameter lifetime distribution with increasing, decreasing and constant hazard rate. This distribution allows the occurrence of zero values and involves the exponential, linear exponential and other combinations of Weibull distributions as submodels. Many statistical properties of the distribution are derived. Maximum likelihood estimation of the parameters and a bias corrective approach is investigated with a simulation study for performance of the estimators. Four real data sets are analyzed for illustrative purposes and it is noted that the distribution is a highly alternative to the gamma, Weibull, Lognormal and exponentiated exponential distributions. [ABSTRACT FROM AUTHOR]

Ground Control Points (GCPs) are needed in most geometric processing of satellite images. Generally used geometric model is based on the use of rational polynomials. The Coefficients of Rational Polynomials (RPCs) are provided by image vendors which are contaminated by some biases. In this paper, a no-GCP bias correction method for RPCs of satellite stereo images is introduced. The method uses global DEMs as ground control information: First, a point cloud is generated from stereo pair, then using the DEM matching strategy it is aligned to the global DEM for estimation of 3D rigid transformation parameters. This transformation is performed with our originally developed method which separates three planimetric parameters from three leveling parameters. They are then employed for bias correction in object space. For DEM matching and also for bias correction, we developed new formulae given directly in geodetic longitude and latitude format, instead of Cartesian map projection coordinates. Numerical results of this research are reported in two categories: with or without (1) GCPs and (2) map projection step. Experiments on two Cartosat-1 stereo pairs show in category (1), improvement in geopositioning accuracy from 399.2 m and 124.0 m to 7.6 m and 2.6 m, respectively. In category (2), we observed RMS improvement in both datasets and in all components up to 3.1 m in longitude, 6.8 m in latitude and 1.2 m in height. [ABSTRACT FROM AUTHOR]

Non-gravitational force models are critical not only for the applications of satellite orbit determination and prediction, but also for the studies of gravitational reference sensors in space-based gravitational wave detection missions and accelerometers in gravity satellite missions. In this paper, based on the inertial sensor data from the TianQin-1 (TQ-1) mission, a correction has been made in the non-gravitational force models by applying additional terms related to the orbital periods. After taking into account this correction, about 37 hours of TQ-1 inertial sensor data is calibrated in the sensitive axes, i.e. y - and z -axes, by comparing with the simulated non-gravitational accelerations. It is indicated that the peak-to-peak value of the non-gravitational acceleration correction terms are about 2% and 13% of the measured accelerations in the y - and z -axes, respectively. Within the frequency band below 0.01 Hz, the root mean square of calibration residual errors in y - and z -axes are suppressed from 1.03 Ă— 10âˆ'9 and 3.872 Ă— 10âˆ'9 m sâˆ'2 to 8.14 Ă— 10âˆ'10 and 1.343 Ă— 10âˆ'9 m sâˆ'2, respectively. The bias and scale factor of the inertial sensor are also obtained from the calibration by the method of least-squares fit. Meanwhile, the inertial sensor measurements are validated and their signal compositions are analyzed. [ABSTRACT FROM AUTHOR]

Objectives: When using variance component estimation to determine the relative weight ratio, the least square solution is generally used as the initial value of the iteration. In geodetic joint inversion, the least squares method will cause ill ⁃ posed problems, so the regularized solution is used instead of the least square solution. Regularization introduces bias to reduce variance, but when using variance component estimation to determine the relative weight ratio, the influence of bias is not considered, and the introduction of bias will cause inaccurate variance component estimation. Methods: This paper adopts the bias - corrected variance component estimation method to eliminate the influence of bias introduced by regularization. The residual - based bias - corrected variance component estimation and the variance component estimation method are used for simulation experiments, the Visso earthquake and Norcia earthquake are used for verification. Results and Conclusions: Simulation experimental result shows that the variance component estimation method after bias correction can better reverse the slip distribution. The bias-corrected variance component estimation method takes into account the bias introduced by the iterative initial value, and the theory is more rigorous. Two real earthquakes results show that the bias correction is reasonable and advantageous. [ABSTRACT FROM AUTHOR]

Under the context of global warming, compound dry and hot events (CDHEs) will increase and bring serious losses to society and the economy. The projection of CDHEs is of great significance for policy-making and risk assessment. In this paper, two large ensemble simulations, CanESM2-LE and CESM-LE, are used to estimate the risk of extreme CDHEs under different warming scenarios in China. First, the biases of the model in the simulation of the temperature and precipitation over the China region are corrected, and the index of CDHEs is established based on a copula function. The results show that extreme CDHEs will occur more often in China with the increase in global warming and the more severe extreme CDHEs are, the greater the risk will be in the future with higher uncertainties. Events that would be attained once every 50 and 100 years in the current climate from CESM-LE (CAanESM2-LE) will be 1.2/1.6 (1.1/1.5) times and 1.3/2.3 (1.5/2.0) times more likely to occur in a 1.5°C/2.0°C warmer climate, respectively. Northwestern China will experience the greatest increase in the risk of extreme CDHEs. Extreme CDHEs expected once every 100 years in the current period over NW China are expected to occur approximately every 5 and 4 years under a 4.0°C warmer world in CanESM2-LE and CESM-LE, respectively. [ABSTRACT FROM AUTHOR]

Feature extraction and classification plays a crucial role in the automated analysis of breast ultrasound (BUS) images. Due to varying sonographic characteristics of benign and malignant lesions, the texture and shape features are mostly employed for designing computer-aided diagnosis (CAD) systems of BUS images. The existing CAD systems use features that are extracted either from the lesion segmented area obtained through segmentation techniques or a rectangular region of interest (ROI) extracted under the guidance of expert Radiologists. However, the significance of features extracted from region comprising only the lesion area is still little explored. This paper investigates the significance of features extracted from the lesion area, lesion surrounding area and rectangular ROI for classification of BUS images. The experiments were conducted on the database of 294 BUS images (104 benign and 190 malignant). Initially, the acquired BUS images were preprocessed through speckle reducing anisotropic diffusion (SRAD) for speckle noise removal. The preprocessed images are segmented using a hybrid segmentation approach including a combination of region-based active contour driven by region-scalable fitting (RBACM-RSF) model and multi-scale Gaussian kernel fuzzy c-means clustering with spatial bias correction (MsGKFCM_S) for getting ROI confined area. The segmented images were further partitioned into two parts (lesion area and lesion surrounding area). Subsequently, a total of 457 texture and shape attributes were extracted from within the lesion area, lesion surrounding area and rectangular ROI comprising of both lesion and its surrounding area. The significance of these features is evaluated using different classifiers (i.e. support vector machine (SVM), Back-propagation artificial neural network (BPANN), Random Forest, AdaBoost). The results indicate that features extracted from within lesion area achieve a maximum classification accuracy of 98.980% with the lowest computational time when linear kernel-based SVM is used. [ABSTRACT FROM AUTHOR]

Our aim at this paper is to estimate the parameters of extended Burr XII (EBXII) based on k-records with corrected bias. To get this goal maximum likelihood estimator (MLE) is applied and bias correction for ML estimation is considered to get better and improved estimator. Numerical results for comparison of performances of estimators are also presented. [ABSTRACT FROM AUTHOR]

The development of engineering disciplines is aided by research into constructing a characteristic index system, particularly when the blockchain technology is used. The key issue is how to choose multi-level evaluation indicators reasonably while also scientifically defining the weights of indicators at all levels. Subjective empowerment methods are insufficient in terms of subjective influence, whereas objective empowerment methods necessitate a large sample size, a practical problem domain, and complex calculation methods. In response to the need for characteristic index system construction and evaluation research, this paper identifies four first-level indicators, eleven second-level indicators, and twenty-one third-level indicators as the main evaluation dimensions including academic achievements, discipline strength, talent training, and international development. The proposed method combines the Fuzzy Analytic Hierarchy Process (AHP), based on triangular fuzzy numbers, with the critic weighting analysis method. It is establishing a multi-level evaluation index system to propose targeted combination weighting methods for the engineering disciplines. To avoid evaluation bias caused by the single use of subjective or objective weighting methods, the difference coefficient method is used for combined weighting based on subjective and objective information to calculate the weighting results. The experimental and modelling data show that the calculation and evaluation results of the algorithm proposed in this paper are promising and applicable in multiple domains where there is a hierarchy in the problem domain and multiple parameters participate in decision making. With the evaluation results of the proposed approach, the subjective weight obtained is 0.047, and the objective weight obtained is 0.253. [ABSTRACT FROM AUTHOR]

The saltmarsh carbon storage potential is a key topic in blue carbon research and climate policy. Ecosystem service valuations provide valuable information to policymakers for habitat management and climate change mitigation policies. Yet, only few saltmarsh valuation studies have included the carbon storage service in the UK context. This paper investigates how the public values saltmarsh ecosystem services, focussing on the carbon storage service. We used a choice experiment to elicit the willingness to pay (WTP) of a representative sample of the Scottish public to support interventions that would maintain or improve the provision of these services. Furthermore, we tested the effect of information on individuals' preferences and WTP with a split sample approach where one group received a treatment in the form of additional information. We found that (i) all attributes had a significant influence on individuals' choices; (ii) both groups had, on average, a positive marginal WTP for all presented ecosystem services; (iii) the treated sample had, on average, no significantly different marginal WTP for carbon storage than the control group. This paper adds to the limited literature on the saltmarsh carbon storage ecosystem service and demonstrates a developed nation's public's openness to nature-based climate change mitigation solutions. [Display omitted] • The Scottish public on average prefers saltmarsh management over the status quo. • Providing additional information reduced heterogeneity in WTP for carbon storage. • Most respondents preferred an improvement of all saltmarsh ecosystem services. • Therefore management should focus on both climate change mitigation and adaptation. [ABSTRACT FROM AUTHOR]

This paper proposes a simple and powerful optimal integration (OPI) method for improving hourly quantitative precipitation forecasts (QPFs, 0-24 h) of a single-model by integrating the benefits of different biascorrected methods using the high-resolution CMA-GD model from the Guangzhou Institute of Tropical and Marine Meteorology of China Meteorological Administration (CMA). Three techniques are used to generate multi-method calibrated members for OPI: deep neural network (DNN), frequency-matching (FM), and optimal threat score (OTS). The results are as follows: (1) The QPF using DNN follows the basic physical patterns of CMA-GD. Despite providing superior improvements for clear-rainy and weak precipitation, DNN cannot improve the predictions for severe precipitation, while OTS can significantly strengthen these predictions. As a result, DNN and OTS are the optimal members to be incorporated into OPI. (2) Our new approach achieves state-of-the-art performances on a single model for all magnitudes of precipitation. Compared with the CMA-GD, OPI improves the TS by 2.5%, 5.4%, 7.8%, 8.3%, and 6.1% for QPFs from clear-rainy to rainstorms in the verification dataset. Moreover, OPI shows good stability in the test dataset. (3) It is also noted that the rainstorm pattern of OPI relies heavily on the original model and that OPI cannot correct for deviations in the location of severe precipitation. Therefore, improvements in predicting severe precipitation using this method should be further realized by improving the numerical model's forecasting capability. [ABSTRACT FROM AUTHOR]

In this paper, time extension methods, originally designed for clear-sky land surface conditions, are used to estimate high-spatial resolution surface daily longwave (LW) radiation from the instantaneous Global LAnd Surface Satellite (GLASS) longwave radiation product. The performance of four time methods were first tested by using ground based flux measurements that were collected from 141 global sites. Combined with the accuracy of daily LW radiation estimated from the instantaneous GLASS LW radiation, the linear sine interpolation method performs better than the other methods and was employed to estimate the daily LW radiation as follows: The bias/Root Mean Square Error (RMSE) of the linear sine interpolation method were −6.30/15.10 W/m2 for the daily longwave upward radiation (LWUP), −1.65/27.63 W/m2 for the daily longwave downward radiation (LWDN), and 4.69/26.42 W/m2 for the daily net longwave radiation (LWNR). We found that the lengths of the diurnal cycle of LW radiation are longer than the durations between sunrise and sunset and we proposed increasing the day length by 1.5 h. The accuracies of daily LW radiation were improved after adjusting the day length. The bias/RMSE were −4.15/13.74 W/m2 for the daily LWUP, −1.3/27.52 W/m2 for the daily LWDN, and 2.85/25.91 W/m2 for the daily LWNR. We are producing long-term surface daily LW radiation values from the GLASS LW radiation product. [ABSTRACT FROM AUTHOR]

Obtaining high-quality quantitative precipitation forecasts is a key precondition for hydrological forecast systems. Due to multisource uncertainties (e.g., initial conditions, model structures, and parameters), raw forecasts are subject to systematic biases; hence, statistical postprocessing is often required to reduce these errors before the forecasts can proceed to hydrological applications. Machine learning (ML) algorithms are canonical statistical models, and they are diverse in type and variation. It is important to verify and compare their performance in the same scenario (e.g., precipitation postprocessing). In this paper, we conduct a large-scale comparison study for the major ML models with diverse model structures and regularization strategies as postprocessors for improving the quality of precipitation forecasts. Specifically, we compare the efficiency and effectiveness of 21 ML algorithms on solving this task. Daily reforecast precipitation with lead times up to 8 days from the Global Ensemble Forecast System and corresponding observations are employed to determine the usability of different models in the Yalong River basin in China. The performance of each model is validated by a group of carefully designed experiments and statistical metrics. The results reveal that improvements in model structures are more effective than regularization strategies. Among these algorithms, the optimized extra-trees regressor exhibits the best performance, effectively reducing overestimation and achieving the best skill in forecasting precipitation. Eleven ensemble members and a 3-day forward-rolling time window can be used as predictors to obtain the best model performance. The systematic experiments and findings also offer useful guidelines for other related studies. [ABSTRACT FROM AUTHOR]

In this paper, we present the new method for real-time GPS receiver bias estimation by using Lagrange interpolation, which is also compared to the two current methods, polynomial and minimization of standard deviation. The estimated method is proposed to reduce the complexity and time of the GPS receiver bias estimation. Lagrange interpolation is the method to find the derivatives and integrals of discrete functions in GPS receiver bias data. The test site is located on Chumphon station, Thailand. The test period of data method is during the year 2004–2019. In the quiet and disturbed days, the polynomial method gives the highest value of the GPS receiver bias at −5.75 ns and −4.25 ns, respectively, but the Lagrange interpolation shows the lowest value of GPS receiver bias at −6.85 ns and −5.25 ns, in order. The results and comparisons among the polynomial GPS receiver bias method, the minimization of standard deviation of GPS receiver bias method, and Lagrange interpolation method show that the calculated time for Lagrange interpolation is shorter compared to other methods and it can be given more time points for finding GPS receiver biases than others. [ABSTRACT FROM AUTHOR]

Breast cancer has high incidences and mortality rates in women worldwide. Malignancy could be detected manually by experienced pathologists based on Hematoxylin and Eosin (H&E) stained images. However, it is time‐consuming and experience‐dependent, making early diagnosis a big challenge. In this paper, a methodology for breast cancer classification based on histopathological images with deep learning was described. A residual learning‐based convolutional neural network named myResNet‐34 was designed for malignancy‐and‐benign classification. In addition, an algorithm automatically generating the target image for stain normalization was proposed, which eliminated the bias caused by manual selection of the reference image. Elastic distortion was introduced and combined with affine transformation for data augmentation considering the characteristics of the H&E images. Experiments were conducted on BreakHis dataset with the proposed framework. Promising results were achieved with an average classification accuracy of around 91% on image‐level classification. Results indicated that both our data augmentation and stain normalization effectively improved the classification accuracy by 2‐3%. [ABSTRACT FROM AUTHOR]

We correct a logic mistake in our paper "On statistical convergence and strong Cesàro convergence by moduli" (León-Saavedra et al. in J. Inequal. Appl. 23:298, 2019). [ABSTRACT FROM AUTHOR]

In traditional unit commitment models, the ramping process of coal-fired generators is often represented by either a static average ramping rate or a once-changing rate model. However, this approach fails to accurately capture the actual ramping process of the unit, resulting in scheduling biases, particularly in systems with renewable energy sources. To address this issue, we propose a dynamic piecewise linear ramping (DPWLR) model that allows arbitrary multi-changes of ramping rates within a single period. We shift our focus from slope changes to time changes. By using a series of location indicator variables that satisfy the type 2 special ordered set, we locate the forward and backward time change limits and then determine the up and down power output limits of the units at each hour. We tested the intraperiod multiple-changing DPWLR model, as part of the unit commitment model, on different systems, including a 3-unit system, a 10-unit system, the IEEE RTS-79, and a 100-unit system. Comparative analysis demonstrates the functionality and superior performance of our proposed model. • A dynamic ramping model to simulate the unit's ramping process is proposed. • The model allows multiple ramping rate changes in one period. • The model uses location indicator variables to identify the ramping limits. • Some reproducibility studies on previous papers are carried out. • Comparison studies between the proposed model and the existing model are conducted. [ABSTRACT FROM AUTHOR]

Underwater active electrosense serves as a valuable complement to conventional underwater target detection technologies, such as acoustic and optics, by offering advantages in short-range and murky environment. However, there is a need to enhance the precision of target location and enrich the research on tracking moving target. This paper delves in to the investigation of locating and tracking underwater sphere target using active electrosense. To improve location accuracy, an approach is proposed to generate a global adaptive model correction coefficient through calibration and general regression neural network (GRNN) fitting. A detection system and sensing array with vertically arranged transmitting and receiving electrodes are designed for experimental research, enabling high-resolution target locating and tracking without the need of any active movement of the array. In the scenario where the target is metallic sphere with a diameter of 50 mm, the results of location experiment indicate that the least square (LS) algorithm achieved 24.7% higher location performance than the multiple signal classification (MUSIC) algorithm under an accurate uniform model correction coefficient, and the location error of the LS algorithm was further reduced by 33.2% using the global adaptive model correction coefficient generated by proposed approach. Based upon these findings, tracking functionality based on discrete location method is achieved under various trajectories and velocities. [Display omitted] • A sensing array with vertically oriented transmitting and receiving electrodes was designed. • The sensing array enables high-resolution of target locating without any active movement. • The performances of the LS and MUSIC algorithms were studied by location experiments. • GRNN is utilized to generate model correction coefficient for location accuracy improving. • Tacking of target were achieved under various of trajectories and velocities. [ABSTRACT FROM AUTHOR]

In this paper we propose a reduced order model (ROM) for incompressible flows based on proper orthogonal decomposition (POD) and having a finite element approximation as full order model (FOM). The main idea is to start from a purely POD-based ROM, projecting the equations onto the ROM space, and then add a nonlinear correction that depends on the ROM unknowns to enhance the final ROM model. This correction is based on the fact that we do have some available high fidelity data, namely, the snapshots. Thus, the correcting term is built as an artificial neural network (ANN) constructed with the snapshots as the training set. This correction is then introduced in the fully discrete ROM system to achieve more accurate solutions. A further feature of our approach is that we construct both the ROM and the FOM using the variational multi-scale (VMS) concept, and this allows us to understand the correction term as an approximation to the scales that are lost when passing from the FOM to the ROM. The resulting corrected ROM has a significant higher accuracy than the original one. [ABSTRACT FROM AUTHOR]

Underwater structure inspection faces great challenges because of the poor measuring environment. To conduct health monitoring for underwater bridge piers, this paper describes a method mainly composed of multicamera digital image correlation, a multirefraction correction model and an approach of grayscale mapping. Verification tests were carried out to examine the accuracy of the reconstruction results both in air and water, and the effectiveness of the multirefraction correction model was also proven. An underwater measuring system was built and was successfully applied to measure the round surface of a concrete pier underwater. The axial measurement error of the points reconstructed after multirefraction modification under the cylindrical coordinate system was only 1.58% of the diameter of the bridge pier. Furthermore, the gray information was transferred to the three-dimensional point cloud. This paper may promote the detection efficiency and accuracy of underwater bridge pier health monitoring. • A shape measurement method for underwater bridge piers is proposed. • A new calibration method of refraction plane is proposed. • 3D topography of the practical bridge pier was realized. [ABSTRACT FROM AUTHOR]

A systematic RF characterization and small-signal parameter extraction for 28 nm fully depleted silicon on insulator (FDSOI) MOSFETs is presented in this paper. Two-step and hybrid de-embedding methodologies have been applied and compared for the accurate de-embedding of the on-chip RF test pads and access interconnect line elements. To avoid the accumulated error from the de-embedding for the small signal model of the FDSOI MOSFET, an error check criterion is adopted and utilized. Then, the complete extrinsic and intrinsic small-signal parameters of the device were extracted at multi-bias points based on the combination of direct extraction and linear regression approaches. For the extrinsic parameter extraction, the substrate-related parasitic elements and the extrinsic resistance components have been extracted. For the intrinsic small-signal parameter extraction, a complete model is presented and applied for direct extraction of the small-signal parameters. To verify the validity of the proposed modeling approach and the employed small-signal model, the S-parameters of a 28 nm FDSOI NMOS device with a 16 × 1 μm gate width were measured and characterized. From the comparison of the measured and modeled data, it is found that the proposed de-embedding and small-signal model provide an accurate characterization of RF parameter extraction for the FDSOI MOSFETs up to 110 GHz. [ABSTRACT FROM AUTHOR]

In this paper, the formation control of the underactuated surface vehicles (USVs) is studied in the presence of the condition of actuator fault and limited communication resources. Considering the performance characteristics of the underactuated surface vehicle (USV), a robust adaptive synchronized event-triggered formation control method is proposed under the usage of the event-triggered control (ETC) and robust neural damping technique. In consideration of the constant energy expenditure of the law of adaptation, the formulated control method is set in the mechanism of triggering by both the controller and the gain related adaptive law, which could cause the communication to reduce more much power consumption. The robust adaptive rule is fault compensated in the way of combining robust neural damping and dynamic surface control (DSC) techniques. Since the fault amendment strategy compresses inconclusive bias fault parameters and model uncertainty parameters, system stability can be enhanced under the premise of reducing the influence of adaptive parameters. A huge volume of work has taken place to ensure semi-global uniform ultimate bounded (SGUUB) stability. To conclude, the effectiveness of the algorithm is verified by two arithmetic examples. • An expanded square parallel search guidance law is proposed using the DVS principle. • A synchronized event-triggered control algorithm is developed for USVs. • The improved fault compensation strategy is designed to stabilize the fault effect. • The ETC algorithm can effectively reduce the burden of communication resources. [ABSTRACT FROM AUTHOR]

Holes in timber beams can significantly affect their bearing capacity. This paper presents an extensive experimental investigation of the effect of circular holes with and without plywood reinforcement on glulam joists behaviour. The tests consider the variability of hole position, number and diameter. In the second step, a finite element model based on fracture mechanics was developed and validated against the experimental force–displacement curves of thirteen configurations. The model reproduces crack initiation and propagation through the adoption of cohesive contact layers. The satisfactory agreement with the experimental data has been the base of extensive parametric analyses considering multiple beam and hole geometry selections and two load arrangements at the upper and lower side of the beam. Finally, the results of the parametric analyses, initially used for a qualitative understanding of the structural behaviour, are used for calibrating probabilistic capacity models of the capacity of simply-supported beams with circular holes. The mechanics-based probabilistic model calculates the capacity as the product between the analytical capacity associated with the reduced cross-section and an adimensional correction factor. The factor is expressed as a linear combination of a set of explanatory variables selected after a step-wise deletion process. • Experimental tests on composite GLT beams with holes. • Effect of holes number, geometry, position and plywood reinforcement. • Validation of a nonlinear finite element model. • Automatic parametric analyses. • Probabilistic capacity model. [ABSTRACT FROM AUTHOR]

Several studies have reported that student evaluation of teaching (SET) presents important problems. First, depending on the area, there are significant differences in the evaluations. Second, numerous noninstructional biases exist, such as when those teachers who award better grades obtain better SETs. Correcting the rankings by considering these biases (e.g., adjusting SETs according to the class grade) has been proposed. In this paper, we analyse a third problem: it is impossible to correct the biases because they are specific to each area, level, and even class. On a sample of 15,439 SETs, we compared the biases present in two very close areas (accounting and finance) and at two levels (undergraduate and postgraduate). Then, we used a procedure based on the analysis of residuals in OLS models to eliminate area- and level-specific biases. However, there are still latent biases apparently linked to each specific group of students. [Display omitted] • Noninstructional biases generate huge distortions in Student Evaluation of Teaching. • These biases are markedly different according to area and level. • A bias correction procedure based on the analysis of residuals in OLS models is proposed. • After incorporating area- and level-specific corrections, there are remaining latent biases. • It does not seem possible to turn SETs into reliable instruments of teaching quality. [ABSTRACT FROM AUTHOR]

In the present paper, a closed-form solution of the Reddy beam theory is developed and applied, for the first time, to investigate the bending behavior of straight and curved functionally graded (FG) beams, where the material properties change continuously from one surface to another in the thickness (or height) direction. The obtained closed-form solution, sum of polynomial and exponential terms, enriches the polynomial displacement field usually proposed in a finite element (FE) approach, with effects also on the derived strain and stress quantities, particularly relevant in FG beams. The adopted beam model is exploited to satisfy parabolic variation of the shear stress distribution along the thickness direction and does not require the use of shear correction factors, particularly difficult to obtain when the beam is inhomogeneous in the thickness direction. Comparative studies are carried-out to establish the robustness and the performance of the present model, and numerical results are presented and discussed in detail to investigate the effects of volume fraction index, radius of curvature, length-to-height ratio, and boundary conditions on the stress response of FG beams. The obtained results can serve as benchmarks for future research. • Closed form solution of Reddy beam model is developed for the first time. • Stress field analysis of Functionally Graded Beam shows the effectiveness of the proposed solution. • A FEM-like approach allows the analysis of beams with complex geometries. [ABSTRACT FROM AUTHOR]

Shot peening is an essential process for forming large thin-walled components in aerospace industries. However, it is challenging to determine directly the mapping relationship between shot peening parameters and deformation response because of the complex elastic–plastic deformation of shot-peened components. In this paper, a computational model of bending deformation of peened plate samples is established by examining the shot peening behavior from the aspect of energy equivalence, in which the deformation energy of the shot-peened plate is determined by the total kinetic energy input due to the shot impacts and a correction factor that takes into account the influences of oblique impact, shot interaction and shot overlap on the peen forming process in practice. It is proven that the square of curvature radius is proportional to the cube of the thickness but inversely proportional to the indentation coverage and the average kinetic energy input of a single shot. For the effect of a single shot impact in peen forming, the average kinetic energy input of a single shot is a power function with the indentation diameter. • Deformation of peened plate is explained from the aspect of energy equivalence. • The square of curvature radius is proportional to the cube of the thickness. • The square of curvature radius is inversely proportional to the coverage. • The square of curvature radius is inversely proportional to the kinetic energy. • Kinetic energy input is a power function with the indentation diameter. [ABSTRACT FROM AUTHOR]

The quality of a data-driven model is heavily dependent on the quality of data. Data from building operation often have data bias problems, which means that the data sample is collected in a way that some members of the intended data population are less likely to be included than others. Data-driven energy forecasting models built on such data hence are biased and could lead to large forecasting errors. Active learning— an effective method to defying data bias—is rarely studied or applied in the area of data-driven building energy forecasting modeling. This paper attempts to fill this gap and explores the application of active learning in data-driven building energy forecasting. The developed strategy in this paper efficiently generate informative training data within a time budget and uses block design to passively consider weather disturbances. The developed active learning strategy is applied and evaluated in both virtual and real-building testbeds against traditional data-driven methods. Via these virtual and real-building evaluation cases, we have demonstrated that the data bias problem typically exists in building operation data is resolved by applying the developed active learning strategy. Building energy forecasting models trained from data generated from the active learning strategy have shown improved performances in both model accuracy and model extendibility perspectives. The effectiveness of the block design module is also validated to effectively consider the impact of weather conditions on active learning design. [ABSTRACT FROM AUTHOR]