Search

Bayesian optimal inference is often heralded as a principled, general framework for human perception. However, optimal inference requires integration over all possible world states, which quickly becomes intractable in complex real-world settings. Additionally, deviations from optimal inference have been observed in human decisions. A number of approximation methods have previously been suggested, such as sampling methods. In this study, we additionally propose point estimate observers, which evaluate only a single best estimate of the world state per response category. We compare the predicted behavior of these model observers to human decisions in five perceptual categorisation tasks. Compared to the Bayesian observer, the point estimate observer loses decisively in one task, ties in two and wins in two tasks. Two sampling observers also improve upon the Bayesian observer, but in a different set of tasks. Thus, none of the existing general observer models appears to fit human perceptual decisions in all situations, but the point estimate observer is competitive with other observer models and may provide another stepping stone for future model development.

Data-driven methods have been extensively utilized in establishing predictive models from historical data for process monitoring and prediction of quality variables. However, most data-driven approaches assume that training data and testing data come from steady-state operating regions and follow the same distribution, which may not be the case when it comes to complex industrial processes. To avoid these restrictive assumptions and account for practical implementation, a novel online transfer learning technique is proposed to dynamically learn cross-domain features based on the variational Bayesian inference in this work. Stemming from the probabilistic slow feature analysis, a transfer slow feature analysis (TSFA) technique is presented to transfer dynamic models learned from different source processes to enhance prediction performance in the target process. In particular, two weighting functions associated with transition and emission equations are introduced and updated dynamically to quantify the transferability from source domains to the target domain at each time instant. Instead of point estimation, a variational Bayesian inference scheme is designed to learn the parameters under probability distributions accounting for corresponding uncertainties. The effectiveness of the proposed technique with applications to soft sensor modelling is demonstrated by a simulation example, a public dataset and an industrial case study.

Models for financial risk often assume that underlying asset returns are stationary. However, there is strong evidence that multivariate financial time series entail changes not only in their within-series dependence structure, but also in the cross-sectional dependence among them. In particular, the stressed Value-at-Risk of a portfolio, a popularly adopted measure of market risk, cannot be gauged adequately unless such structural breaks are taken into account in its estimation. A method for consistent detection of multiple change points in high-dimensional panel data set is proposed where both conditional variance of individual time series and their correlations are allowed to change over time. The consistency of the proposed method in multiple change point estimation is proved, and its good performance is demonstrated through simulation studies and an application to the Value-at-Risk problem on a real dataset. The change point detection methodology is implemented in the R package segMGarch , available from CRAN.

Motivated by the need for effectively summarising, modelling, and forecasting the distributional characteristics of intra-daily returns, as well as the recent work on forecasting histogram-valued time-series in the area of symbolic data analysis, we develop a time-series model for forecasting quantile-function-valued (QF-valued) daily summaries for intra-daily returns. We call this model the dynamic quantile function (DQF) model. Instead of a histogram, we propose to use a $g$-and-$h$ quantile function to summarise the distribution of intra-daily returns. We work with a Bayesian formulation of the DQF model in order to make statistical inference while accounting for parameter uncertainty; an efficient MCMC algorithm is developed for sampling-based posterior inference. Using ten international market indices and approximately 2,000 days of out-of-sample data from each market, the performance of the DQF model compares favourably, in terms of forecasting VaR of intra-daily returns, against the interval-valued and histogram-valued time-series models. Additionally, we demonstrate that the QF-valued forecasts can be used to forecast VaR measures at the daily timescale via a simple quantile regression model on daily returns (QR-DQF). In certain markets, the resulting QR-DQF model is able to provide competitive VaR forecasts for daily returns., Comment: MATLAB code: https://github.com/wilson-ye-chen/aqua

This article presents a framework to determine driving style and design a driver steering model considering driver characteristics. First, principal component analysis (PCA) and K-means clustering are utilized to classify 30 participants into cautious, moderate, and aggressive drivers. Subsequently, a generic steering model is established based on the model predictive control method. Thereafter, the maximum lateral acceleration is extracted as a crucial indicator to represent driver characteristics, and it is calibrated through probabilistic models using the dataset, which consists of the classified drivers. Besides, point estimation model and interval estimation model are leveraged to determine driving style and adjust constraints in the stochastic programming-based steering model. Finally, simulation experiments present the variations of actual output trajectories between the aggressive drivers and the cautious drivers.

Rail surface defect inspection based on machine vision faces challenges against the complex background with interference and severe data imbalance. To meet these challenges, we regard defect detection as a key-point estimation problem and present the attention neural network for rail surface defect detection via CASIoU-guided center-point estimation (CCEANN). CCEANN contains two crucial components. One is the stacked attention Hourglass backbone via cross-stage fusion of multi-scale features (CSFA-Hourglass), in which the convolutional block attention module with variable receptive fields (VRF-CBAM) is introduced, and a two-stage Hourglass structure balancing the network depth and feature fusion plays a key role. Furthermore, the CASIoU-guided center-point estimation head module (CASIoU-CEHM) integrating the delicate coordinate compensation mechanism regresses detection boxes flexibly to adapt to defects' large-scale variation, in which the proposed CASIoU loss, a loss regressing the consistency of Intersection-over-Union (IoU), central-point distance, area ratio, and scale ratio between the targeted defect and the predicted defect, achieves higher regression accuracy than state-of-the-art IoU-based losses. The experiments demonstrate that CCEANN outperforms competitive deep learning-based methods in four surface defect datasets.

Purpose Mokken scale analysis (MSA) is an attractive scaling procedure for ordinal data. MSA is frequently used in health-related quality of life research. Two of MSA's prime features are the scalability coefficients and the automated item selection procedure (AISP). The AISP partitions a (large) set of items into scales based on the observed item scores; the resulting scales can be used as measurement instruments. There exist two issues in MSA: First, point estimates, standard errors, and test statistics for scalability coefficients are inappropriate for clustered item scores, which are omnipresent in quality of life research data. Second, the AISP insufficiently takes sampling fluctuation of Mokken’s scalability coefficients into account. Methods We solved both issues by providing point estimates and standard errors for the scalability coefficients for clustered data and by implementing a Wald-based significance test in the AISP algorithm, resulting in a test-guided AISP (T-AISP), that is available for both nonclustered and clustered test scores. Results We integrated the T-AISP into a two-step, test-guided MSA for scale construction, to guide the analysis for nonclustered and clustered data. The first step is performing a T-AISP and select the final scale( s ). For clustered data, within-group dependency is investigated on the final scale(s). In the second step, the strength of the scale(s) is determined and further analyses are performed. The procedure was demonstrated on clustered item scores obtained from administering a questionnaire on quality of life in schools to 639 students nested in 30 classrooms. Conclusions We developed a two-step, test-guided MSA for scale construction that takes into account sample fluctuation of all scalability coefficients and that can be applied to item scores obtained by a nonclustered or clustered sampling design.

The limited sample size in multi-stress accelerated life testing makes the conventional large-sample-based inference methods inefficient. To overcome this problem, this paper develops a new statistical inference method based on random variable transformation for multi-stress accelerated life testing with Weibull distribution and progressive Type-II censoring. Firstly, a χ2 statistic is constructed using the random variable transformation method, and the exact point estimates of model parameters are derived based on the χ2 statistic. Subsequently, based on the χ2 statistic, the exact confidence interval of shape parameter is provided, while the generalized confidence intervals of the accelerated model parameters are calculated by constructing the new multivariate generalized pivotal quantities. Furthermore, the generalized confidence intervals of some commonly used reliability indexes are provided for better reliability management and decision making. Finally, a simulation study and a real case study are conducted to illustrate the implementation and effectiveness of the proposed method.

1.AbstractPedigree inference from genotype data is a challenging problem, particularly when pedigrees are sparsely sampled and individuals may be distantly related to their closest genotyped relatives. We present a new method that infers small pedigrees of close relatives and then assembles them into larger pedigrees. To assemble large pedigrees, we introduce several new formulas and tools including a new likelihood for the degree separating two small pedigrees, a method for detecting individuals who share background identity-by-descent (IBD) that does not reflect recent common ancestry, and a method for identifying the ancestral branches through which distant relatives are connected. Our method also takes several new approaches that help to improve the accuracy and efficiency of pedigree inference. In particular, we incorporate age information directly into the likelihood rather than using ages only for consistency checks and we employ a heuristic branch-and-bound-like approach to more efficiently explore the space of possible pedigrees. Together, these approaches make it possible to construct large pedigrees that are challenging or intractable for current inference methods. The new method, Bonsai, is available at https://github.com/23andMe/bonsaitree.

Under partial shading conditions (PSCs), a series-connected photovoltaic (PV) string often shows multi-peak power-voltage (P-V) characteristics, which indicates a multi-modal distribution of maximum power points (MPPs). The phenomenon brings challenges to the maximum power point estimation (MPPE), since a single model usually cannot perform well in different modes. To address this problem, a real-time shading identification process is proposed to identify the shading information of the applied PV string. The measured data are divided into several fields according to the obtained number of peaks in the P-V curve. Multiple Gaussian process regression (M-GPR) models are then proposed to predict the MPP locus in a specific data field. Experiments are conducted to evaluate the accuracy of the proposed method. Simulation and experimental results show that the shading information enables the proposed M-GPR to obtain more accurate estimation performance compared to its single-model counterparts.

The literature review highlighted the impacts of drivers’ behavior on passengers’ attitudes in the choice of seatbelt usage. However, limited studies have been done to determine those impacts. Studying the passengers’ seatbelt use is especially needed to find out why passengers choose not to buckle up, and consequently it helps decision makers to target appropriate groups. So, this study was conducted to find drivers’ characteristics that might impact the passenger’s seatbelt use, in addition to other passengers’ characteristics themselves. While performing any analysis, it is important to use a right statistical model to achieve a less biased point estimate of the model parameters. The latent class multinomial logit model (LC-MNL) can be seen as an alternative to the mixed logit model, replacing the continuous with a discrete distribution, by capturing possible heterogeneity through membership in various clusters. In this study, instead of a response to the survey or crash observations, we employed a real-life observational data for the analysis. Results derived from the analysis reveal a clear indication of heterogeneity across individuals for almost all parameters. Various socio-demographic variables for class allocation and models with different latent numbers were considered and checked in terms of goodness of fit. The results indicated that a class membership with three factors based on vehicle type would result in a best fit. The results also highlighted the significant impacts of driver seatbelt status, time of a day, distance of traveling, vehicle type, and driver gender, instead of passenger gender, as some of the factors impacting the passengers’ choice of seatbelt usage. In addition, it was found that the belting status of passengers is positively associated with the belting condition of drivers, highlighting the psychological behavioral impact of drivers on passengers. Extensive discussion has been made regarding the implications of the findings.

Adaptation of a classifier to new domains is one of the challenging problems in machine learning. This has been addressed using many deep and non-deep learning based methods. Among the methodologies used, that of adversarial learning is widely applied to solve many deep learning problems along with domain adaptation. These methods are based on a discriminator that ensures source and target distributions are close. However, here we suggest that rather than using a point estimate obtaining by a single discriminator, it would be useful if a distribution based on ensembles of discriminators could be used to bridge this gap. This could be achieved using multiple classifiers or using traditional ensemble methods. In contrast, we suggest that a Monte Carlo dropout based ensemble discriminator could suffice to obtain the distribution based discriminator. Specifically, we propose a curriculum based dropout discriminator that gradually increases the variance of the sample based distribution and the corresponding reverse gradients are used to align the source and target feature representations. An ensemble of discriminators helps the model to learn the data distribution efficiently. It also provides a better gradient estimates to train the feature extractor. The detailed results and thorough ablation analysis show that our model outperforms state-of-the-art results., Comment: This work is an extension of our BMVC-2019 paper (arXiv:1907.10628)

Risk management is considered as a vital process contributing to the successful outcome of a complex construction project in terms of achieving the associated project objectives. The widely used industrial practice in managing construction project risks is to assign probability and impact values to each risk and to map risks on a risk matrix. The main criticism of this practice relates to ignoring complex interdependencies between risks and using point estimates for probability and impact values. Furthermore, risks mapped on a matrix are deemed to influence a specific objective and there is a challenge involved in aggregating the impact of risks across multiple (conflicting) project objectives. Utilizing a data-driven Bayesian Belief Network methodology, in this paper we introduce a new process where the risks mapped on a risk matrix corresponding to each project objective are aggregated and modeled as a risk network, and a holistic impact of each risk is captured across the network by means of new risk metrics. The proposed methodology is demonstrated through a real application. The results specific to the two ranking schemes (assuming independence/interdependence of risks) are found to be negatively correlated, which substantiates the importance of utilizing an interdependency-based risk management process.

PurposeThe primary objective of this study is to propose a robust multiobjective solution search approach for a mean-variance multiple correlated quality characteristics optimisation problem, so-called “multiple response optimisation (MRO) problem”. The solution approach needs to consider response surface (RS) model parameter uncertainties, response uncertainties, process setting sensitivity and response correlation strength to derive the robust solutions iteratively.Design/methodology/approachThis study adopts a new multiobjective solution search approach to determine robust solutions for a typical mean-variance MRO formulation. A fine-tuned, non-dominated sorting genetic algorithm-II (NSGA-II) is used to derive efficient multiobjective solutions for varied mean-variance MRO problems. The iterative search considers RS model uncertainties, process setting uncertainties and response correlation structure to derive efficient fronts. The final solutions are ranked based on two different multi-criteria decision-making (MCDM) techniques.FindingsFive different mean-variance MRO cases are selected from the literature to verify the efficacy of the proposed solution approach. Results derived from the proposed solution approach are compared and contrasted with the best solution(s) derived from other approaches suggested in the literature. Comparative results indicate significant superiorities of the top-ranked predicted robust solutions in nondominated frequency, closeness-to-target and response variabilities.Research limitations/implicationsThe solution approach depends on RS modelling and considers continuous search space.Practical implicationsIn this study, promising robust solutions are expected to be more suitable for implementation than point estimate-based MOO solutions for a real-life MRO problem.Originality/valueNo evidence of earlier research demonstrates the superiority of a MOO-based iterative solution search approach for mean-variance MRO problems by simultaneously considering model uncertainties, response correlation and process setting sensitivity.

Null hypothesis significance testing is the major statistical procedure in fMRI, but provides only a rather limited picture of the effects in a data set. When sample size and power is low relying only on strict significance testing may lead to a host of false negative findings. In contrast, with very large data sets virtually every voxel might become significant. It is thus desirable to complement significance testing with procedures like inferiority and equivalence tests that allow to formally compare effect sizes within and between data sets and offer novel approaches to obtain insight into fMRI data. The major component of these tests are estimates of standardized effect sizes and their confidence intervals. Here, we show how Hedges' g, the bias corrected version of Cohen's d, and its confidence interval can be obtained from SPM t maps. We then demonstrate how these values can be used to evaluate whether nonsignificant effects are really statistically smaller than significant effects to obtain “regions of undecidability” within a data set, and to test for the replicability and lateralization of effects. This method allows the analysis of fMRI data beyond point estimates enabling researchers to take measurement uncertainty into account when interpreting their findings., Inferiority and equivalence tests allow to compare effect sizes within and between data sets and are based on standardized effect sizes and their confidence intervals. We show how Hedges'g and its confidence interval can be obtained from SPM t maps and demonstrate how these can be used to obtain “regions of undecidability” within a data set, to test for the replicability, and for the lateralization of effects.

We describe the application of Bayesian hierarchical models to the analysis of data from long-term, environmental monitoring programs. The goal of these ongoing programs is to understand status and trend in natural resources. Data are usually collected using complex sampling designs including stratification, revisit schedules, finite populations, unequal probabilities of inclusion of sample units, and censored observations. Complex designs intentionally create data that are missing from the complete data that could theoretically be obtained. This “missingness” cannot be ignored in analysis. Data collected by monitoring programs have traditionally been analyzed using the design-based Horvitz–Thompson estimator to obtain point estimates of means and variances over time. However, Horvitz–Thompson point estimates are not capable of supporting inference on temporal trend or the predictor variables that might explain trend, which instead requires model-based inference. The key to applying model-based inference to data arising from complex designs is to include information about the sampling design in the analysis. The statistical concept of ignorability provides a theoretical foundation for meeting this requirement. We show how Bayesian hierarchical models provide a general framework supporting inference on status and trend using data from the National Park Service Inventory and Monitoring Program as examples. Supplemental Materials Code and data for implementing the analyses described here can be accessed here: https://doi.org/10.36967/code-2287025.

In many biomedical experiments, such as toxicology and pharmacological dose–response studies, one primary goal is to identify a threshold value such as the minimum effective dose. This paper applies Fisher’s fiducial idea to develop an inference method for these threshold values. In addition to providing point estimates, this method also offers confidence intervals. Another appealing feature of the proposed method is that it allows the use of multiple parametric relationships to model the underlying pattern of the data and hence, reduces the risk of model mis-specification. All these parametric relationships satisfy the qualitative assumption that the response and dosage relationship is monotonic after the threshold value. In practice, this assumption may not be valid but is commonly used in dose–response studies. The empirical performance of the proposed method is illustrated with synthetic experiments and real data applications. When comparing to existing methods in the literature, the proposed method produces superior results in most synthetic experiments and real data sets. Supplementary materials accompanying this paper appear on-line.

A real time scenario of dynamic primary user (PU) is considered in additive Laplacian noise. Two transitions or status changes of PU in the fixed sensing time are considered. The last status change point (LSCP) is estimated with maximum likelihood estimation by using dynamic programming. We consider Cumulative Sum (CuSum) based weighted samples for detection. We consider three detection schemes such as sample mean detector, energy detection and improved absolute value cumulation detection. We derive closed form expressions of detection probability $$(P_D)$$ and false alarm probability $$(P_F)$$ for all the three schemes. We present our results with receiver operating characteristic (ROC) for the considered schemes. We also present simulation results, which are closely matching with their analytical counterparts. We compare the ROC of the considered system with the ROC of conventional techniques. In the conventional techniques, all the samples in the sensing time are used for detection without LSCP estimation and weight. It is found that the considered system outperforms the conventional schemes.

We present a new way to estimate the lifetime distribution of a reparable system consisted of similar (equal) components. We consider as a reparable system, a system where we can replace a failed component by a new one. Assuming that the lifetime distribution of all components (originals and replaced ones) are the same, the position of a single component can be represented as a renewal process. There is a considerable amount of works related to estimation methods for this kind of problem. However, the data has information only about the time of replacement. It was not recorded which component was replaced. That is, the replacement data are available in an aggregate form. Using both Bayesian and a maximum likelihood function approaches, we propose an estimation procedure for the lifetime distribution of components in a repairable system with aggregate data. Based on a latent variables method, our proposed method out-perform the commonly used estimators for this problem. The proposed procedure is generic and can be used with any lifetime probability model. Aside from point estimates, interval estimates are presented for both approaches. The performances of the proposed methods are illustrated through several simulated data, and their efficiency and applicability are shown based on the so-called cylinder problem. The computational implementation is available in the R package srplv.

Network autocorrelation models (NAMs) are widely used to study a response variable of interest among subjects embedded within a network. Although the NAM is highly useful for studying such networked observational units, several simulation studies have raised concerns about point estimation. Specifically, these studies have consistently demonstrated a negative bias of maximum likelihood estimators (MLEs) of the network effect parameter. However, in order to gain a practical understanding of point estimation in the NAM, these findings need to be expanded in three important ways. First, these simulation studies are based on relatively simple network generative models rather than observed networks, thereby leaving as an open question how realistic network topologies may affect point estimation in practice. Second, although there has been strong work done in developing two-stage least squares estimators as well as Bayesian estimators, only the MLE has received extensive attention in the literature, thus leaving practitioners in question as to best practices. Third, the performance of these estimators need to be compared using both bias and variance, as well as the coverage rate of each estimator's corresponding confidence or credible interval. In this paper we describe a simulation study which aims to overcome these shortcomings in the following way. We first fit real social networks using the exponential random graph model and used the Bayesian predictive posterior distribution to generate networks with realistic topologies. We then compared the performance of the three different estimators mentioned above.

In the process of building electrical load data collection, it is inevitable to introduce different kinds of noises, which makes the observation values deviate from the actual values, thus resulting in high levels of uncertainties. And such uncertainties make it difficult to achieve accurate point prediction of the short-term building electrical load. To improve the rationality of the prediction results and offer more effective information for decision makers, this paper proposes a novel multi-objective algorithm optimized modular fuzzy method which can accomplish the interval prediction for the short-term electrical load. First, one novel single-input-rule-modules (SIRMs)-based distributed interval fuzzy model (SIRM-DIFM) is proposed by replacing the original functional weights of the traditional SIRMs-based fuzzy inference system (SIRM-FIS) with the interval functional weights. Then, a data-driven learning scheme is presented for constructing the SIRM-DIFM. This learning sheme includes two main steps. The first step utilizes the iterative least square method to generate fuzzy rules for the SIRMs and determine the centers of the interval functional weights, while in the second step, the genetic algorithm (GA)-based multi-objective optimization algorithm is adopted to determine the widths of the interval functional weights. Through these two steps, accurate point estimation and reasonable interval prediction results can be achieved. Finally, two building electrical load prediction experiments are conducted to verify the effectiveness of the presented SIRM-DIFM. Simulation results indicate that the proposed SIRM-DIFM can compensate the shortcomings of the low accuracy of the point estimation and the predicted interval can effectively cover the observed data, providing the decision-makers more reliable and useful information.

Quite frequently, catastrophes impact populated areas of the world, and hence the need for proper risk evaluations that support mitigation and management processes. Because of the uncertain nature of extreme natural hazards and lack of data, forecasts of the potential damage and losses before the event happens are needed. Catastrophe (CAT) models build on scenarios that represent all possible realizations of the hazard in terms of recurrence and intensity. Probabilistic risk models require the characterization of the hazards, the exposure model for the infrastructure, and its vulnerability. The main objective of this research is to compare loss exceedance curves, probable maximum loss curves, and average annual losses using four different available seismic hazard models for Chile. To isolate the effect of changing the hazard model in the risk results, the exposure and vulnerability information is fixed to the one available from the Global Assessment Report, GAR 15, and GAR ATLAS 2017. Imprecise probability theory, logic trees, and frequency and severity blends used by CAT modelers are the approaches applied and compared herein to propose either model blending or an interval of possible realizations. Both types of results have pros and cons. Blended results are point estimates which make them useful in a more traditional way, but their computation necessarily implies assigning weights to the models according to the modeler preferences. On the other hand, raw intervals of variability (without any knowledge of how the variable is distributed inside) are more transparent, as they simply state the bound of what is known without any preference, but their use is less understood among practitioners and could be even impractical.

To make valid statistical inferences from mediation analysis, a number of assumptions need to be assessed. Among the assumptions, 2 frequently discussed ones are (a) the independent variable, mediator, and outcome variables are measured without error; and (b) no confounders of the effects in the mediation model are omitted. The impact of violating either assumption alone on statistical inference of mediation has been discussed in previous literature. In practice, violations of the 2 assumptions often co-occur. In this study, we analytically investigated the effects of measurement error and omitting confounders on statistical inference of mediation effects, including both point estimation and significance testing. Based on the analytical results, we proposed sensitivity analysis techniques for assessing the robustness of mediation inference to the violation of the 2 assumptions. To implement the techniques, we developed R functions and a user-friendly web tool. Simulated-data and real-data examples were provided for illustrations. We hope the developed tools will help researchers conduct sensitivity analyses of mediation inference more conveniently. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

Methanol to olefin (MTO) technology provides the opportunity to produce olefins from nonpetroleum sources such as coal, biomass and natural gas. More than 20 commercial MTO plants have been put into operation. Till now, contributions on optimal operation of industrial MTO plants from a process systems engineering perspective are rare. Based on relevance vector machine (RVM), a data-driven framework for optimal operation of the industrial MTO process is established to fully utilize the plentiful industrial data sets. RVM correlates the yield distribution prediction of main products and the operation conditions. These correlations then serve as the constraints for the multi-objective optimization model to pursue the optimal operation of the plant. Nondominated sorting genetic algorithm II is used to solve the optimization problem. Comprehensive tests demonstrate that the ethylene yield is effectively improved based on the proposed framework. Since RVM does provide the distribution prediction instead of point estimation, the established model is expected to provide guidance for actual production operations under uncertainty.

Change point estimation in standard process observed over time is an important problem in literature with applications in various fields. We study this problem in a heterogeneous population. A model-based clustering procedure relying on skewed matrix-variate mixture is proposed. It is capable of capturing the heterogeneity pattern and estimating change points from all data groups simultaneously. The appeal of such approach also lies in its flexibility to model the skewness and dependence in data with good interpretability. Two novel algorithms called matrix power mixture with abrupt change model and matrix power mixture with gradual change model are developed. The approaches are illustrated by simulation studies across a variety of settings. The models are then tested on the US crime data with promising results.

Deep learning is quickly becoming essential to human ecosystem. However, the opacity of certain deep learning models poses a legal barrier in its adoption for greater purposes. Explainable AI (XAI) is a recent paradigm intended to tackle this issue. It explains the prediction mechanism produced by black box AI models, making it extremely practical for safety, security or financially important decision making. In another aspect, most deep learning studies are based on point estimate prediction with no measure of uncertainty which is vital for decision making. Obviously, these works are not suitable for real world applications. This paper presents a Remaining Useful Life (RUL) estimation problem for turbofan engines equipped with prognostic explainability and uncertainty quantification. A single input, multi outputs probabilistic Long Short-Term Memory (LSTM) is employed to predict the RULs distribution of the turbofans and SHapley Additive exPlanations (SHAP) approach is applied to explain the prognostic made. The explainable probabilistic LSTM is thus able to express its confidence in predicting and explains the produced estimation. The performance of the proposed method is comparable to several other published works

Binary outcomes are often encountered when analyzing cluster randomized trials (CRTs). A common approach to obtaining the average treatment effect of an intervention may involve using a logistic regression model. We outline some interpretive and statistical challenges associated with using logistic regression and discuss two alternative/supplementary approaches for analyzing clustered data with binary outcomes: the linear probability model (LPM) and the modified Poisson regression model. In our simulation and applied example, all models use a standard error adjustment that is effective even if a low number of clusters is present. Simulation results show that both the LPM and modified Poisson regression models can provide unbiased point estimates with acceptable coverage and type I error rates even with as little as 20 clusters.

The power law process (PLP) (i.e., the nonhomogeneous Poisson process with power intensity law) is perhaps the most widely used model for analyzing failure data from reliability growth studies. Statistical infer- ences and prediction analyses for the PLP with left-truncated data with classical methods were extensively studied by Yu et al. (2008) recently. However, the topics discussed in Yu et al. (2008) only included maximum likelihood estimates and condence intervals for parameters of interest, hy- pothesis testing and goodness-of-t test. In addition, the prediction limits of future failure times for failure-truncated case were also discussed. In this paper, with Bayesian method we consider seven totally dierent prediciton issues besides point estimates and prediction limits for xn+k. Specically, we develop estimation and prediction methods for the PLP in the presence of left-truncated data by using the Bayesian method. Bayesian point and credible interval estimates for the parameters of interest are derived. We show how ve single-sample and three two-sample issues are addressed by the proposed Bayesian method. Two real examples from an engine develop- ment program and a repairable system are used to illustrate the proposed methodologies.

Background\ud \ud Within the UK, standard methods of economic evaluation centre on the maximisation of the quality-adjusted life-year (QALY). However, preference-based measures used to estimate QALYs may not be suited to all economic evaluations, as they routinely measure only health outcomes.\ud \ud \ud \ud Aim\ud \ud This study used an economic evaluation alongside a clinical trial (EEACT) comparing patients’ preferences for a telephone versus a face-to-face consultation to incorporate process utility into cost-effectiveness analyses.\ud \ud \ud \ud Methods\ud \ud An EEACT is described that generates QALYs using Short-form 6-dimension version 1 (SF-6Dv1) responses. These results exclude specific consideration of process utility. A health state valuation study is then reported that bolts a process domain onto the SF-6Dv1 using data obtained from the EEACT. These results therefore include the consideration of process utility. The results of the EEACT with and without process utility are then compared.\ud \ud \ud \ud Results\ud \ud This study shows that the QALY, in its current form, does not capture patient benefits associated with the process of receiving healthcare. The EEACT illustrates this, showing a statistically significant difference between control and intervention groups for the patient experience questionnaire communication domain, indicating that patients preferred the intervention. This preference was not identified in the cost-effectiveness outcomes, and the point estimates lie in the north-west quadrant of the cost-effectiveness plane. The preference is captured after adding a communication domain. The point estimate moves to the north-east quadrant, where the intervention is more effective and more costly than the control.\ud \ud \ud \ud Conclusion\ud \ud This study indicates that it is possible to capture patients’ preferences for processes associated with care, in a format compatible with the QALY.

With the development of deep learning, object detection algorithms based on horizontal box are widely used in the field of damage identification. However, damages can be in any direction and position, and they are not necessarily horizontal or vertical. This article proposes a bolt damage identification network, namely, orientation-aware center point estimation network, which models a damage as a center point of its rotated bounding box. The proposed orientation-aware center point estimation network uses deep layer aggregation network to search center points and regress to all other damage properties, such as size and angle. A loss function is designed to improve the optimization efficiency of network. Orientation-aware center point estimation network is applied to bolt damage detection, and comparison with the well-known Faster Region-Convolutional Neural Network (a benchmark using horizontal bounding box) demonstrates the accuracy of the proposed method. Finally, videos were utilized to verify the capability of the proposed orientation-aware center point estimation network in real-time detection of bolt damages.

Accurate travel time prediction is very important for real-time traveller information systems. Many existing traveller information systems provide point estimates of forecast travel times. Often th...

Hospitals commonly project demand for their services by combining their historical share of regional demand with forecasts of total regional demand. Hospital-specific forecasts of demand that provide prediction intervals, rather than point estimates, may facilitate better managerial decisions, especially when demand overage and underage are associated with high, asymmetric costs. Regional forecasts of patient demand are commonly available as a Poisson random variable, e.g., for the number of people requiring hospitalization due to an epidemic such as COVID-19. However, even in this common setting, no probabilistic, consistent, computationally tractable forecast is available for the fraction of patients in a region that a particular institution should expect. We introduce such a forecast, DICE (Demand Intervals from Consistent Estimators). We describe its development and deployment at an academic medical center in California during the `second wave' of COVID-19 in the Unite States. We show that DICE is consistent under mild assumptions and suitable for use with perfect, biased, unbiased regional forecasts. We evaluate its performance on empirical data from a large academic medical center as well as on synthetic data.

Capacity plays a crucial role in a port’s competitive position and the growth of its market share. An investment decision to provide new port capacity should be supported by a growing demand for port services. However, port demand is volatile and uncertain in an increasingly competitive market environment. Also, forecasting models themselves are associated with epistemic uncertainty due to model and parameter uncertainties. This paper applies a Bayesian statistical method to forecast the annual throughput of the multipurpose Port of Isafjordur in Iceland. Model uncertainties are thus taken into account, while parameter uncertainties are handled by selecting influencing macroeconomic variables based on mutual information analysis. The presented model has an adaptive capability as new information becomes available. Our method results in a range of port throughput forecasts, in addition to a point estimate, and it also accounts for epistemic uncertainty, thus increasing the reliability of forecasts. Our results provide support for informed decision-making in capacity planning and management. Our forecasts show a constant linear growth of containerized throughput the period 2020–2025. Noncontainerized throughput declines rapidly over the same period.

Publisher Copyright: CCBY Calibrating stochastic radio channel models to new measurement data is challenging when the likelihood function is intractable. The standard approach to this problem involves sophisticated algorithms for extraction and clustering of multipath components, following which, point estimates of the model parameters can be obtained using specialized estimators. We propose a likelihood-free calibration method using approximate Bayesian computation. The method is based on the maximum mean discrepancy, which is a notion of distance between probability distributions. Our method not only by-passes the need to implement any high-resolution or clustering algorithm, but is also automatic in that it does not require any additional input or manual pre-processing from the user. It also has the advantage of returning an entire posterior distribution on the value of the parameters, rather than a simple point estimate. We evaluate the performance of the proposed method by fitting two different stochastic channel models, namely the Saleh-Valenzuela model and the propagation graph model, to both simulated and measured data. The proposed method is able to estimate the parameters of both the models accurately in simulations, as well as when applied to 60 GHz indoor measurement data.

Meta-analysis aggregates quantitative outcomes from multiple scientific studies to produce comparable effect sizes. The resultant integration of useful information leads to a statistical estimate with higher power and more reliable point estimate when compared to the measure derived from any individual study. Effect sizes are usually estimated using mean differences of the outcomes of treatment and control groups in experimental studies. Although different software exists for the calculations in meta-analysis, understanding how the calculations are done can be useful to many researchers, particularly where the values reported in the literature data is not applicable in the software available to the researcher. In this paper, search was conducted online primarily using Google and PubMed to retrieve relevant articles on the different methods of calculating the effect sizes and the associated confidence intervals, effect size correlation, p values and I2, and how to evaluate heterogeneity and publication bias are presented.

Buildings consume over 40% of the total energy in modern societies, and improving their energy efficiency can significantly reduce our energy footprint. In this paper, we present \texttt{WattScale}, a data-driven approach to identify the least energy-efficient buildings from a large population of buildings in a city or a region. Unlike previous methods such as least-squares that use point estimates, \texttt{WattScale} uses Bayesian inference to capture the stochasticity in the daily energy usage by estimating the distribution of parameters that affect a building. Further, it compares them with similar homes in a given population. \texttt{WattScale} also incorporates a fault detection algorithm to identify the underlying causes of energy inefficiency. We validate our approach using ground truth data from different geographical locations, which showcases its applicability in various settings. \texttt{WattScale} has two execution modes -- (i) individual, and (ii) region-based, which we highlight using two case studies. For the individual execution mode, we present results from a city containing >10,000 buildings and show that more than half of the buildings are inefficient in one way or another indicating a significant potential from energy improvement measures. Additionally, we provide probable cause of inefficiency and find that 41\%, 23.73\%, and 0.51\% homes have poor building envelope, heating, and cooling system faults, respectively. For the region-based execution mode, we show that \texttt{WattScale} can be extended to millions of homes in the US due to the recent availability of representative energy datasets., This paper appeared in the Journal ACM Transactions on Data Science

For better performance of hyperspectral target detectors, the prior target spectrum is expected to be accurate and consistent with the target spectrum in the hyperspectral image to be detected. The existing hyperspectral target detection algorithms usually assume that the prior target spectrum is highly reliable. However, the label obtained is not always precise in practice, and pixels of the same object may have quite different spectra. Since it is hard to acquire a highly reliable prior target spectrum in some application scenarios, we propose a Bayesian constrained energy minimization (B-CEM) method for hyperspectral target detection. Instead of the point estimation of the target spectrum, we infer the posterior distribution of the true target spectrum based on the prior target spectrum. Specifically, considering the characteristics of hyperspectral image and target detection task, we adopt the Dirichlet distribution to approximate the true target spectrum. Experimental results on three datasets demonstrate the effectiveness of the proposed B-CEM when the known target spectrum is noisy or inconsistent with the true target spectrum. The necessity to approximate the true target spectrum is also proved. Generally, the distributional estimate achieves better performance than using the known target spectrum directly.

There has been increasing interest in modeling survival data using deep learning methods in medical research. In this paper, we proposed a Bayesian hierarchical deep neural networks model for modeling and prediction of survival data. Compared with previously studied methods, the new proposal can provide not only point estimate of survival probability but also quantification of the corresponding uncertainty, which can be of crucial importance in predictive modeling and subsequent decision making. The favorable statistical properties of point and uncertainty estimates were demonstrated by simulation studies and real data analysis. The Python code implementing the proposed approach was provided.

Distributed Bayesian inference provides a full quantification of uncertainty offering numerous advantages over point estimates that autonomous sensor networks are able to exploit. However, fully-decentralized Bayesian inference often requires large communication overheads and low network latency, resources that are not typically available in practical applications. In this paper, we propose a decentralized Bayesian inference approach based on stochastic gradient Langevin dynamics, which produces full posterior distributions at each of the nodes with significantly lower communication overhead. We provide analytical results on convergence of the proposed distributed algorithm to the centralized posterior, under typical network constraints. We also provide extensive simulation results to demonstrate the validity of the proposed approach.

Lately, researchers have widely centered around renewable energy assets, for example, wind control and solar units to decrease the utilization of fossil fuels as the fundamental wellspring of ecological natural contaminations. One of the fundamental difficulties for the generation of wind and solar energies is that they frequently including the uncertainty because of the stochastic natures of wind speeds and solar radiation. Along these lines, the current uncertainty to assets the wind and solar units and necessary to assess the arranging strategy of distribution systems for having a dependable execution. To display the uncertainty related along with the wind and solar power, the point estimate method (PEM) is utilized. Weibull and Beta distributions are utilized to deal with uncertain information factors. These fundamental goals present work is to minimize the generation cost of the framework is enhanced dependent on the butterfly optimization algorithm (BOA). Four case test frameworks viewed as it is discovered that the proposed strategy gives better arrangement as far as execution time and normal cost viability. The recreation results demonstrate that the entrance of sustainable energy source builds, the generation cost diminishes. The outcomes acquired along with the butterfly optimization algorithm contrasted and another one understood strategy. Also, the precise distribution of generation cost.

Many reliability analysis methods require complicated mathematical process or access to comprehensive datasets. Such shortcomings limit their application to solve the geotechnical problems. It is of advantage to develop simpler reliability analysis methods that can be employed in the design of geotechnical structures. The current study suggests a simple framework for quick reliability analysis of deep excavation projects in urban areas, which is a common geotechnical problem. To investigate the feasibility of the presented method, five case studies were considered. It is worth mentioning that the horizontal displacement at the crest of excavation was set to be the main system response. For verification purposes, the results were compared to the random set, point estimate and Monte Carlo methods results, which are also used for reliability analysis of geotechnical problems. All case studies were recognized as projects of high importance and monitored during the excavation process. The field observations confirmed that the estimated probabilities of excessive deformations were reasonable for all cases. Comparing the modeling results and field measurements suggests the feasibility of the presented method for evaluating the reliability of deep urban excavations and estimating the horizontal displacement at the crest of excavations.

Expert-coded datasets provide scholars with otherwise unavailable cross-national longitudinal data on important concepts. However, expert coders vary in their reliability and scale perception, potentially resulting in substantial measurement error; this variation may correlate with outcomes of interest, biasing results in analyses that use these data. This latter concern is particularly acute for key concepts in peace research. In this article, I describe potential sources of expert error, focusing on the measurement of identity-based discrimination. I then use expert-coded data on identity-based discrimination to examine: 1) the implications of measurement error for quantitative analyses that use expert-coded data, and 2) the degree to which different techniques for aggregating these data ameliorate these issues. To do so, I simulate data with different forms and levels of expert error and regress conflict onset on different aggregations of these data. These analyses yield two important results. First, almost all aggregations show a positive relationship between identity-based discrimination and conflict onset consistently across simulations, in line with the assumed true relationship between the concept and outcome. Second, different aggregation techniques vary in their substantive robustness beyond directionality. A structural equation model provides the most consistently robust estimates, while both the point estimates from an Item Response Theory (IRT) model and the average over expert codings provide similar and relatively robust estimates in most simulations. The median over expert codings and a naive multiple imputation technique yield the least robust estimates.

With complicated structures and complex physicochemical reactions, most industrial processes are intrinsically characterized by high dimensionality, nonlinearity, non-Gaussianity, and self-correlation. In this article, a novel probabilistic generative model, called weighted copula regression (WCR), is developed for complex processes. This method employs a probabilistic graphical tool (vine copula) to flexibly handle the underlying patterns via the factorization of complex dependence structures into multiple bivariate copulas. Monte Carlo estimation is incorporated for establishing a fast and reliable computing framework. To avoid overinterpretation of the industrial data, an adaptive sample selection strategy is proposed to explore the underlying distribution of the sample space and to select the most “informative” samples. By considering the copula weights that carry crucial information on the local data, the probabilistic WCR method can provide fast point estimate with prediction uncertainty for every predicted sample. The proposed WCR method is compared with six state-of-the-art methods, and the efficiency is validated using a numerical example and the ethylene cracking furnace process.

Perkembangan pemanfaatan panel surya di masa depan akan terus meningkat. Salah satu bentuk karakteristik panel surya merupakan kurva I-V yang mana dengan kurva tersebut dapat digunakan untuk menganalisa besaran daya keluaran panel surya. Dengan mengetahui kurva I-V tersebut dapat dilakukan Maximum Power Point Estimation (MPPE) yang dapat diampu oleh panel surya. Informasi mengenai nilai estimasi daya maksimum panel surya merupakan bagian penting untuk menentukan kapasitas pembebanan, selain itu juga untuk menjaga umur peralatan yang digunakan. Feed Forward Neural Network dengan Algoritma Back Propagation (FFBP) terbukti dapat memberikan informasi nilai MPPE pada keluaran panel surya. Nilai masukan pada ANN berupa tegangan dan arus dari panel surya, sedangkan keluaran dari ANN tersebut berupa nilai estimasi daya. Hasil dari simulasi MPPE didapatkan galat rata rata sebesar 0.04 poin antara daya aktual (MPP) dan daya estimasi (MPPE).

In this work, we utilize a multi-column CNNs framework to estimate the gaze point of a person sitting in front of a display from an RGB-D image of the person. Given that gaze points are determined by head poses, eyeball poses, and 3D eye positions, we propose to infer the three components separately and then integrate them for gaze point estimation. The captured depth images, however, usually contain noises and black holes which prevent us from acquiring reliable head pose and 3D eye position estimation. Therefore, we propose to refine the raw depth for 68 facial keypoints by first estimating their relative depths from RGB face images, which along with the captured raw depths are then used to solve the absolute depth for all facial keypoints through global optimization. The refined depths will provide us reliable estimation for both head pose and 3D eye position. Given that existing publicly available RGB-D gaze tracking datasets are small, we also build a new dataset for training and validating our method. To the best of our knowledge, it is the largest RGB-D gaze tracking dataset in terms of the number of participants. Comprehensive experiments demonstrate that our method outperforms existing methods by a large margin on both our dataset and the Eyediap dataset.

Procedural material models have been gaining traction in many applications thanks to their flexibility, compactness, and easy editability. We explore the inverse rendering problem of procedural material parameter estimation from photographs, presenting a unified view of the problem in a Bayesian framework. In addition to computing point estimates of the parameters by optimization, our framework uses a Markov Chain Monte Carlo approach to sample the space of plausible material parameters, providing a collection of plausible matches that a user can choose from, and efficiently handling both discrete and continuous model parameters. To demonstrate the effectiveness of our framework, we fit procedural models of a range of materials---wall plaster, leather, wood, anisotropic brushed metals and layered metallic paints---to both synthetic and real target images., 12 pages, 13 figures, Pacific Graphics 2020

This work presents an application of upgraded arithmetic operations on trapezoidal fuzzy number to extend the reliability from point estimation to interval estimation. A realistic example of an acc...

Imaging of multiple maritime moving targets is a challenging task in the synthetic aperture radar (SAR) system owing to the fact that the complex target motion produces evident image defocusing. Given that the sparsity of ship targets in the SAR image, the authors propose a new imaging method of multiple maritime moving targets based on sparsity Bayesian learning. To avoid the overcomplete velocity dictionary with heavy computational burden, the Gaussian chirplet transform is exploited and improved based on the velocity constraint to estimate target Doppler parameters for the observation matrix construction. An observation model is established for imaging multiple ship targets. In the proposed method, the multiple ship target imaging task is formulated into the sparsity Bayesian framework, which provides a posterior density function for the target image and improves the imaging quality over the conventional methods based on the point estimate. The Bayesian compressive sensing (BCS) using a hierarchical form of the Laplace prior is applied to reconstruct the moving target image. Since BCS provides an estimation of the uncertainty in the reconstruction, the sea clutter can be well suppressed while the multiple targets are refocused. Simulations and experimental Gaofen-3 data are performed to verify the effectiveness of the proposed method.