Search

This paper proposed a prediction algorithm for the degraded actuator taking into account the impact of estimation error of hidden index in the closed-loop system. To this end, a unified prediction framework is established to evaluate the hidden degradation information and recursively update the degradation model parameters simultaneously. The advantage is that the prediction framework can comprehensively compensate the estimation error of hidden degradation index caused by system uncertainty. To jointly estimate the degradation information in avoidance of the impact of system uncertainty, a modified adaptive Kalman filter is designed, and the proof of stability is provided. With the priori estimate from the filter, the degradation model parameters are updated by the inverse filtering probability based on Bayes' theorem. It is followed by the computation of the remaining useful life (RUL) prediction utilizing aforementioned hidden degradation information and the latest degradation model. The effectiveness of the proposed RUL prediction algorithm is demonstrated by the degraded actuator in the continuous casting process. • A novel prediction framework is designed for degraded actuator in closed-loop system. • Adaptive degradation model is established by modified adaptive Kalman filter. • The estimation error of hidden degradation index is integrated. • The stability of optimized adaptive Kalman filter is proven. • Recursive prediction algorithm is performed without full life data. [ABSTRACT FROM AUTHOR]

In robust parameter design for blocked experiments, the correlation of response observations within each block and model parameter uncertainty often impact acquiring ideal operating conditions. In this paper, a Bayesian mixed regression-based multi-response surface modeling and optimization method is suggested to address the above issues. Firstly, the mixed effects model is incorporated into the Bayesian framework, and posterior distributions of the model parameters are derived using Bayes' theorem. Secondly, the hybrid Monte Carlo algorithm is employed to calculate the model parameters. Thirdly, the expected quality loss function satisfying the specification is constructed to lessen the impact of outliers on the results of optimization, and the optimal factor settings are obtained by the hybrid genetic algorithm. In addition, the posterior probability is used to assess the conformance of the optimization results. Finally, a simulated study and real-world example of the additive manufacturing process are used to illustrate the viability of the proposed method. Compared with the current techniques, the proposed method can reduce the impact of model uncertainty on the modeling and optimization results, leading to more conformant and robust optimization results. • The correlation of observations within block and model parameter uncertainty are considered. • The priors of covariance of random effect and standard deviation of random error are extended to non-conjugate priors. • Compared with MCMC, the HMC algorithm improves efficiency. • In the blocked experiment, both quality loss and the conformance of optimization results are simultaneously considered. • On the application side, the problem of quality design for multiple 3D printers printing simultaneously is solved. [ABSTRACT FROM AUTHOR]

In this paper, we study pseudo-labelling. Pseudo-labelling employs raw inferences on unlabelled data as pseudo-labels for self-training. We elucidate the empirical successes of pseudo-labelling by establishing a link between this technique and the Expectation Maximisation algorithm. Through this, we realise that the original pseudo-labelling serves as an empirical estimation of its more comprehensive underlying formulation. Following this insight, we present a full generalisation of pseudo-labels under Bayes' theorem, termed Bayesian Pseudo Labels. Subsequently, we introduce a variational approach to generate these Bayesian Pseudo Labels, involving the learning of a threshold to automatically select high-quality pseudo labels. In the remainder of the paper, we showcase the applications of pseudo-labelling and its generalised form, Bayesian Pseudo-Labelling, in the semi-supervised segmentation of medical images. Specifically, we focus on: (1) 3D binary segmentation of lung vessels from CT volumes; (2) 2D multi-class segmentation of brain tumours from MRI volumes; (3) 3D binary segmentation of whole brain tumours from MRI volumes; and (4) 3D binary segmentation of prostate from MRI volumes. We further demonstrate that pseudo-labels can enhance the robustness of the learned representations. The code is released in the following GitHub repository: https://github.com/moucheng2017/EMSSL. [ABSTRACT FROM AUTHOR]

• Estimating the geographical distribution of species has long been done in Ecology. • Appropriately measuring and showing uncertainty in the models is crucial. • In this paper we provide a method to map uncertainty using cartograms under a Bayesian framework. Predicting the geographical distribution of a species is a central topic in ecology, conservation and management of natural resources especially for invasive organisms. Invasive species can modify the structure and function of invaded ecosystems, altering their biodiversity, and causing significant economic losses locally and globally. Therefore, measuring and visualizing the uncertainty inherent in species' potential distributions is fundamental for effective biodiversity monitoring and planning conservation interventions. This paper discusses a new Bayesian approach to mapping this uncertainty using cartograms, previously published knowledge, and presence/absence data. [ABSTRACT FROM AUTHOR]

Abstract The Expected Hypervolume Improvement (EHVI) is a frequently used infill criterion in Multi-Objective Bayesian Global Optimization (MOBGO), due to its good ability to lead the exploration. Recently, the computational complexity of EHVI calculation is reduced to O (n log n) for both 2-D and 3-D cases. However, the optimizer in MOBGO still requires a significant amount of time, because the calculation of EHVI is carried out in each iteration and usually tens of thousands of the EHVI calculations are required. This paper derives a formula for the Expected Hypervolume Improvement Gradient (EHVIG) and proposes an efficient algorithm to calculate EHVIG. The new criterion (EHVIG) is utilized by two different strategies to improve the efficiency of the optimizer discussed in this paper. Firstly, it enables gradient ascent methods to be used in MOBGO. Moreover, since the EHVIG of an optimal solution should be a zero vector, it can be regarded as a stopping criterion in global optimization, e.g., in Evolution Strategies. Empirical experiments are performed on seven benchmark problems. The experimental results show that the second proposed strategy, using EHVIG as a stopping criterion for local search, can outperform the normal MOBGO on problems where the optimal solutions are located in the interior of the search space. For the ZDT series test problems, EHVIG still can perform better when gradient projection is applied. [ABSTRACT FROM AUTHOR]

General arrangement and distributed system design is a complex problem that is a fundamental aspect of ship design. Current approaches to this design problem employ a paradigm of using automated tools to generate and analyze potential vessel solutions. These approaches rely on the generation and optimization of vessel models based on design parameters. Created vessel models are then evaluated and compared to understand how parameters influence possible vessel characteristics. This process is time and resource intensive, which limits its application in early-stage design when many critical arrangement and distributed system design decisions are made. In this paper a new approach is proposed to complement the automated tool-based paradigm. For a vessel with a defined set of systems to be arranged and connected, the approach measures the probabilistic arrangement and distributed system configuration, without generating vessel models. This efficiently provides leading indicators of the expected design outcomes and resultant vessel characteristics, which can help guide early-stage decisions and lead to better applications of resource-intensive design tools. In this paper, methods supporting this approach are presented and application is demonstrated on a naval frigate concept design. [ABSTRACT FROM AUTHOR]

Automatic text summarization tools help users in the biomedical domain to acquire their intended information from various textual resources more efficiently. Some of biomedical text summarization systems put the basis of their sentence selection approach on the frequency of concepts extracted from the input text. However, it seems that exploring other measures rather than the raw frequency for identifying valuable contents within an input document, or considering correlations existing between concepts, may be more useful for this type of summarization. In this paper, we describe a Bayesian summarization method for biomedical text documents. The Bayesian summarizer initially maps the input text to the Unified Medical Language System (UMLS) concepts; then it selects the important ones to be used as classification features. We introduce six different feature selection approaches to identify the most important concepts of the text and select the most informative contents according to the distribution of these concepts. We show that with the use of an appropriate feature selection approach, the Bayesian summarizer can improve the performance of biomedical summarization. Using the Recall-Oriented Understudy for Gisting Evaluation (ROUGE) toolkit, we perform extensive evaluations on a corpus of scientific papers in the biomedical domain. The results show that when the Bayesian summarizer utilizes the feature selection methods that do not use the raw frequency, it can outperform the biomedical summarizers that rely on the frequency of concepts, domain-independent and baseline methods. [ABSTRACT FROM AUTHOR]

This paper presents an Empirical Bayes before-after evaluation of the road safety effects of a new motorway (freeway) in Østfold county, Norway. The before-period was 1996–2002. The after-period was 2009–2015. The road was rebuilt from an undivided two-lane road into a divided four-lane road. The number of killed or seriously injured road users was reduced by 75 percent, controlling for (downward) long-term trends and regression-to-the-mean (statistically significant at the 5 percent level; recorded numbers 71 before, 11 after). There were small changes in the number of injury accidents (185 before, 123 after; net effect −3%) and the number of slightly injured road users (403 before 279 after; net effect +5%). Motorways appear to mainly reduce injury severity, not the number of accidents. The paper discusses challenges in implementing the Empirical Bayes design when less than ideal data are available. [ABSTRACT FROM AUTHOR]

In this paper the Bayesian methodology, known as Approximate Bayesian Computation (ABC), is applied to the problem of the atmospheric contamination source identification. The algorithm input data are on-line arriving concentrations of the released substance registered by the distributed sensors network. This paper presents the Sequential ABC algorithm in detail and tests its efficiency in estimation of probabilistic distributions of atmospheric release parameters of a mobile contamination source. The developed algorithms are tested using the data from Over-Land Atmospheric Diffusion (OLAD) field tracer experiment. The paper demonstrates estimation of seven parameters characterizing the contamination source, i.e.: contamination source starting position (x,y), the direction of the motion of the source (d), its velocity (v), release rate (q), start time of release (ts) and its duration (td). The online-arriving new concentrations dynamically update the probability distributions of search parameters. The atmospheric dispersion Second-order Closure Integrated PUFF (SCIPUFF) Model is used as the forward model to predict the concentrations at the sensors locations. [ABSTRACT FROM AUTHOR]

In the diversity of contemporary decision-making tasks, where the data is no longer static and changes over time, data stream processing has become an important issue in the field of pattern recognition. In addition, most of the real problems are not balanced, representing their classes in various improportions. Following paper proposes the Prior Imbalance Compensation method, modifying on-the-fly predictions made by the base classifier, aiming at mapping prior probability in the statistics of assigned classes. It is intended to be a less computationally complex competition for popular algorithms such as smote , solving this problem by oversampling the training set. The proposed method has been tested using computer experiments on the example of a set of various data streams, leading to promising results, suggesting its usefulness in solving this type of problems. [ABSTRACT FROM AUTHOR]

• Introducing dynamic classifier selection to outlier detection. • An adaptive KNN rule based on SVDD is proposed to determine validation set. • Using posterior probabilities of classifiers to estimate competences. • Results implying the superiority of dynamic model over single model and static models. Ensembles of outlier detectors are drawing increasing attentions recently, in spite of the difficulty on developing ensembles in the framework of unsupervised learning. We have noted that existing outlier ensembles often use certain fusion rules (e.g. majority voting) to aggregate individual learners. Theoretically, these individuals are assumed to be error-independent so that single models can be outperformed by the ensemble. However, it is of great difficulty to satisfy this assumption in practical applications. By dynamic selecting more competent individual(s) for each test pattern, this problem can be alleviated effectively. Inspired by this idea, this paper proposes a dynamic ensemble outlier detection model using one-class classifiers as base learners. As the competences of base detectors are estimated totally on data points in the validation set, its impact on the selection is significant. In order to achieve an efficient selection, we propose an adaptive k-nearest neighbor (KNN) rule, instead of traditional KNN algorithm, to constitute the validation set for each test pattern. Our adaptive KNN rule firstly uses algorithm support vector data description (SVDD) to mine the local area where class conditional probabilities are not constant in terms of the corresponding test pattern. Competences estimated with neighbor patterns in this area should thus be more accurate than that by traditional KNN rule. A probabilistic model that uses posterior probabilities of one-class classifiers is used then to estimate classifier competences. We present experimental evidence of the detection performance improvement over single models and over a variety of static ensemble models, by using data sets from UCI repository. [ABSTRACT FROM AUTHOR]

• Bayesian inference based state estimation of voltage dips in power system. • Network uncertainties have been incorporated. • A iterative sampling process is used to accelerate the Monte Carlo integration. • The required measurement set-up is quantified. The magnitudes of voltage dips on non-monitored buses can be estimated by applying Bayes's theorem with measured voltage dips on a limited number of monitored buses. Bayes's theorem is used to obtain the grid conditions, including the network impedance, load, distributed generation vector, and short-circuit location, from which the voltage magnitudes along the feeder can be calculated. In order to implement this approach effectively, it is necessary to understand which parameters among the grid conditions are more significant for the estimation performance and how the uncertainties associated with these parameters affect the estimation results. Moreover, since the measurement structure provides observations to calculate the posterior distribution from the prior distribution of the grid conditions. The measurement parameters, quantities, and accuracy levels are varied in simulations to obtain an optimum measurement structure balancing between the cost and the measurement performance. This paper presented the results of the sensitivity studies which have been done on the IEEE 13-Bus and 123-Bus test systems. The results shows the performance of the voltage dip state estimation approach is less sensitive to the uncertainty level of the load while more sensitive to those of the DG, the short-circuit power, the zero sequence impedance of the cable and the short-circuit resistance. The voltage measurement accuracy is more important than that of current and phase angles. The proposed method shows much stronger robustness compared to the conventional one and the estimation performance can be improved through several measures, e.g. adding more measurement points. [ABSTRACT FROM AUTHOR]

Risk analysis of offshore wells decommissioning, and abandonment processes is challenging due to limited life-cycle information of the well, and failure data of safety barriers in place. To this end, it is essential to capture and implement the variability associated with the sparse data for conducting risk analysis with considerable confidence level. The hierarchical Bayesian analysis provides a viable alternative to address the uncertainty of the data through aggregation for each causation. Bayesian network, through its robust computation engine, is used to define dependence of causations and uses Bayes' theorem to update the analysis as new information becomes available. In addition, the Bayesian network helps to represent complex dependencies among causations through appropriate relaxation strategy to minimize uncertainty in the data, link parameter of interest, and overall accident scenario modelling. This paper presents the integration of Hierarchical Bayesian model with a Bayesian network to conduct the risk analysis of well decommissioning and abandonment processes. The proposed methodology is illustrated using a well plugging and abandonment operational failure reported by the Department of Mineral Management Service (MMS). The results demonstrate the potential of the proposed approach as a robust means to study complex well decommissioning activities. [ABSTRACT FROM AUTHOR]

In this paper, we investigate image reconstruction from the Electrical Impedance Tomography (EIT) problem using a statistical inversion method based on Bayes' theorem and an Iteratively Regularized Gauss Newton (IRGN) method. We compare the traditional IRGN method with a new Pilot Adaptive Metropolis algorithm that (i) enforces smoothing constraints and (ii) incorporates a sparse prior. The statistical algorithm reduces the reconstruction error in terms of ℓ 2 and ℓ 1 norm in comparison to the IRGN method for the synthetic EIT reconstructions presented here. However, there is a trade-off between the reduced computational cost of the deterministic method and the higher resolution of the statistical algorithm. We bridge the gap between these two approaches by using the IRGN method to provide a more informed initial guess to the statistical algorithm. Our coupling procedure improves convergence speed and image resolvability of the proposed statistical algorithm. [ABSTRACT FROM AUTHOR]

• Real world problems often involve scarce, uncertain, and heterogeneous data. • Proposed methodology uses Bayesian Statistics to produce diagnoses using such data as input. • Aleatory and Epistemic uncertainties are considered and naturally propagated from source to result. Instrumentation and Control systems are often assumed to break and give no readings under certain conditions, but working perfectly otherwise. In reality, aleatory and epistemic factors create a grey area where operators are often unsure of the validity of sensor measurements. Through the use of Bayes' Theorem, this paper proposes a novel approach that first characterizes both aleatory and epistemic uncertainty, and then combines all available information in a Bayesian network, in order to produce quantitative estimates of unobservable variables in the system. Uncertainties are also propagated from sources to results in a natural manner. The approach was applied to a test case, and was able to identify a Vessel Break transient with quantitative probabilities in a timely manner despite the information being scarce, uncertain, and heterogeneous. The approach was thus demonstrated to be a possible alternative method for decision-making under such non-ideal conditions. [ABSTRACT FROM AUTHOR]

• This paper proposes an optimal sensor configuration strategy targeting at model parameter identification under unknown seismic ground motion. • To form the cost function for sensor configuration, the direct challenge that both the structural model parameters and the seismic time-history input are unknown is effectively overcome in this paper. • Regarding the mandatory constraints in applying GA, a variable conversion strategy of sensor configuration vector from continuous variable representation to binary string representation is also proposed. • The results obtained by the optimal sensor placement method proposed in the paper is validated by experimental case studies. The number of sensors and their position is very important for the acquisition of high-quality measurement data in structural health monitoring (SHM), where the detection and evaluation of structural health status under unknown seismic ground motion have been an important research area. It is undoubtedly that the identification of structural model parameters under unknown seismic ground motion or joint identification of model parameters and seismic time-history input depends on the effective placement of sensors on structures. At present, the investigation of sensor configuration is mainly intended for structural model or modal parameter identification without taking into account the seismic effect, which cannot guarantee the effective contribution of the dynamic measurement data to the joint identification of ground motion and model parameters. To the best of authors knowledge, the optimal sensor configuration for the purpose of model parameter identification under unknown seismic ground motion has not yet been reported in the current literature. In the framework of Bayes' theorem and information theory, this paper proposes an optimal sensor placement methodology targeting at model parameter identification under unknown seismic ground motion. Based on the joint identification of ground motion and structural model parameters, the optimal configuration scheme for a given number of sensors is determined by maximizing a proposed expected entropy measure formed by utilizing the marginal distribution of model parameters prior to the availability of the measurement data. In addition, to solve the combinatorial optimization problem arising from the optimal sensor configuration, this paper also proposes a simple yet efficient strategy for revising the fitness function of genetic algorithm, which converts the continuous variable representation of the sensor configuration vector into a binary string representation. This completely avoids the common approaches relying on modifying various genetic operators to ensure the mandatory constraint of always configuring a fixed number of sensors. The validity of the proposed methodology is fully validated by numerical and experimental case studies conducted respectively for a 20-story shear-building model and a laboratory 6-story shear-building model under shaking table test. [ABSTRACT FROM AUTHOR]

• A trivariate statistical framework is proposed for flood frequency analysis. • It uses a combination of a storm generator and a distributed hydrological model. • The trivariate methodology considers the main flood-generating mechanisms. • Initial soil moisture and spatio-temporal variability of storms are explicit. • This methodology can be replicated thanks its processes-based orientation. This paper proposes a trivariate methodology for flood frequency estimation. It combines the flood peak, storm magnitude, and initial soil moisture condition (ISMC) as the main flood-related statistical variables to be considered. The semi-arid Mediterranean "Rambla del Poyo" catchment has been used as a representative case study where the influence of the spatio-temporal variability of the storms and the ISMC on floods can lead to differences of up to two orders of magnitude in quantiles when the most commonly used methods are applied. In order to incorporate the main flood-generating mechanisms, the integrated use of a multidimensional storm generator with distributed hydrological modelling is proposed. Flood quantiles are then estimated by combining the maximum flows with the storm magnitude and ISMC in a trivariate probability distribution function through the application of Bayes' theorem and Lagrange's Mean Value theorem. Although the methodology proposed in this paper has been applied and tested in only one case study, it can be extended to other case studies due to its process-based orientation. [ABSTRACT FROM AUTHOR]

Numerous papers in the literature recommend the use of Bayes' theorem to merge the available a priori knowledge about a measurand with the measurement result associated to the measurand itself. The obtained a posteriori value should be used in conformity assessment, rather than the measured one, to reduce the risk of erroneous decisions. The assumption behind Bayes' theorem is that the distributions of both the a priori knowledge and the measurements are correct and truly reliable. This means that no deviations or systematic errors in either the manufacturing process or the measuring instrument are considered. The main question is hence whether these assumptions are satisfied or not in the practical industrial cases. And, if they are not satisfied, as it often occurs in the industrial practice how can deviations and systematic errors be considered when applying Bayes' theorem? This paper proposes a solution based on a modified Bayes' theorem using Random-Fuzzy Variables (RFVs) and possibility theory. Different simulations and an experimental case are reported showing the effectiveness of the modified Bayes' theorem in the context of industrial metrology. • Measuring instrument or manufacturing process may deviate in a practical scenario. • Standard Bayes' theorem is not effective or reliable in conformity analysis. • Modified Bayes' theorem using RFVs to include the possible deviations. • Better results in conformity analysis obtained using the Modified Bayes' theorem. [ABSTRACT FROM AUTHOR]

• Error model considers pump delays, pump flow rate, and loading concentration. • Kernel density estimation of error model. • Bayesian uncertainty quantification based on Markov chain Monte Carlo simulations. • Code and case studies published as open source code. Current mechanistic chromatography process modeling methods lack the ability to account for the impact of experimental errors beyond detector noise (e.g. pump delays and variable feed composition) on the uncertainty in calibrated model parameters and the resulting model-predicted chromatograms. This paper presents an uncertainty quantification method that addresses this limitation by determining the probability distribution of parameters in calibrated models, taking into consideration multiple realistic sources of experimental error. The method, which is based on Bayes' theorem and utilizes Markov chain Monte Carlo with an ensemble sampler, is demonstrated to be robust and extensible using synthetic and industrial data. The corresponding software is freely available as open-source code at https://github.com/modsim/CADET-Match. [ABSTRACT FROM AUTHOR]

This paper examines how testing accuracy plays a key role in detecting someone is positive. Bayesian-theorem exploits how an inaccurate rapid antigen test with a high true-positive rate and a low false-positive rate affects the detection of the probability that someone is positive. The probability is calculated with three determinants in actual dental practices: Infection rate, true positive rate, and false positive rate. The result suggests that in a high infection rate, the false positive rate of rapid antigen tests plays a key role in detecting positive individuals. In 6.5% infection rate, 0.991 true positive rate, and 0.0005 false positive rate, the probability is 0.992 while it with 0.05 false positive rate is reduced to 0.579. The proposed Bayesian analysis can be used for future analysis with imprecise tests in other applications. [ABSTRACT FROM AUTHOR]

This paper presents an efficient approach to compute the bounds on the reliability of a structure subjected to uncertain parameters described by means of imprecise probabilities. These imprecise probabilities arise from epistemic uncertainty in the definition of the hyper-parameters of a set of random variables that describe aleatory uncertainty in some of the structure's properties. Typically, such calculation involves the solution of a so-called double-loop problem, where a crisp reliability problem is repeatedly solved to determine which realization of the epistemic uncertainties yields the worst or best case with respect to structural safety. The approach in this paper aims at decoupling this double loop by virtue of the Augmented Space Integral. The core idea of the method is to infer a functional relationship between the epistemically uncertain hyper-parameters and the probability of failure. Then, this functional relationship can be used to determine the best and worst case behavior with respect to the probability of failure. Three case studies are included to illustrate the effectiveness and efficiency of the developed methods. • Augmented reliability formulation for coping with aleatoric and epistemic uncertainty. • Bayes' theorem allows tracking dependency with respect to epistemic uncertainty. • Implementation considers different simulation methods. [ABSTRACT FROM AUTHOR]

The failure-probability-based global sensitivity measure can detect the effect of input variables on the structural failure probability, which can provide useful information in reliability-based design. In this paper, a new efficient simulation method is proposed to estimate the failure-probability-based global sensitivity measure. The proposed method is based on the Bayes' theorem and importance sampling Markov chain simulation. The Bayes' theorem is used to provide a single-loop simulation method and the importance sampling Markov chain simulation is used to further reduce the computational cost. Compared to the traditional double-loop Monte Carlo simulation method, the proposed method requires only a single set of samples to estimate the failure-probability-based global sensitivity measure and its computational cost does not depend on the dimensionality of input variables. Finally, one numerical example and two engineering examples are presented to illustrate the accuracy and efficiency of the proposed method. [ABSTRACT FROM AUTHOR]

Attribute weight assignment plays a key role in multiple attribute decision making (MADM). For the issue of labeled multiple attribute decision making (LMADM), the existing methods of attribute weight determination that have been well developed for MADM usually ignore or do not take full advantage of the supervisory function of labels. As a result, the weights produced by these methods may not be ideal in practice. To make up for this deficiency, this paper develops an objective method based on Bayes risk. Specifically, the LMADM problem is first put forward, then a Gaussian kernel based loss function is proposed to cope with the drawback that the loss function in Bayes risk is usually determined by experts. Meanwhile, Mahalanobis distance and fuzzy neighborhood relationship are employed to measure the fuzziness of data set. Finally, a number of experiments, including the comparison experiments on UCI data and the effectiveness evaluation of fighter, are carried out to illustrate the superiority and applicability of the proposed method. [ABSTRACT FROM AUTHOR]

This paper presents an artificial cognitive system tightly integrating object perception and manipulation for assistive robotics. This is necessary for assistive robots, not only to perform manipulation tasks in a reasonable amount of time and in an appropriate manner, but also to robustly adapt to new environments by handling new objects. In particular, this system includes perception capabilities that allow robots to incrementally learn object categories from the set of accumulated experiences and reason about how to perform complex tasks. To achieve these goals, it is critical to detect, track and recognize objects in the environment as well as to conceptualize experiences and learn novel object categories in an open-ended manner, based on human–robot interaction. Interaction capabilities were developed to enable human users to teach new object categories and instruct the robot to perform complex tasks. A naive Bayes learning approach with a Bag-of-Words object representation are used to acquire and refine object category models. Perceptual memory is used to store object experiences, feature dictionary and object category models. Working memory is employed to support communication purposes between the different modules of the architecture. A reactive planning approach is used to carry out complex tasks. To examine the performance of the proposed architecture, a quantitative evaluation and a qualitative analysis are carried out. Experimental results show that the proposed system is able to interact with human users, learn new object categories over time, as well as perform complex tasks. [ABSTRACT FROM AUTHOR]

Software defect prediction plays a significant part in identifying the most defect-prone modules before software testing. Quite a number of researchers have made great efforts to improve prediction accuracy. However, the problem of insufficient historical data available for within- or cross- project still remains unresolved. Further, it is common practice to use the probability density function for a normal distribution in Naïve Bayes (NB) classifier. Nevertheless, after performing a Kolmogorov–Smirnov test, we find that the 21 main software metrics are not normally distributed at the 5% significance level. Therefore, this paper proposes a new Bayes classifier, which evolves NB classifier with non-normal information diffusion function, to help solve the problem of lacking appropriate training data for new projects. We conduct three experiments on 34 data sets obtained from 10 open source projects, using only 10%, 6.67%, 5%, 3.33% and 2% of the total data for training, respectively. Four well-known classification algorithms are also included for comparison, namely Logistic Regression, Naïve Bayes, Random Tree and Support Vector Machine. All experimental results demonstrate the efficiency and practicability of the new classifier. [ABSTRACT FROM AUTHOR]

In contrast to the situation in a laboratory, the study of the solar atmosphere has to be pursued without direct access to the physical conditions of interest. Information is therefore incomplete and uncertain and inference methods need to be employed to diagnose the physical conditions and processes. One of such methods, solar atmospheric seismology, makes use of observed and theoretically predicted properties of waves to infer plasma and magnetic field properties. A recent development in solar atmospheric seismology consists in the use of inversion and model comparison methods based on Bayesian analysis. In this paper, the philosophy and methodology of Bayesian analysis are first explained. Then, we provide an account of what has been achieved so far from the application of these techniques to solar atmospheric seismology and a prospect of possible future extensions. [ABSTRACT FROM AUTHOR]

This paper presents model parameter estimation conducted by data assimilation and associated uncertainty quantification for predictive engineering with specific application to reflood phenomena in PWR rod bundles. The uncertainties in the prediction of engineering systems are known to be originated from various non-input parameters, e.g., numerics, scaling effects, etc., as well as modeling parameters such as initial and boundary conditions, and physical models. Since the physical models are usually developed by small scale experiments and the experiments used for validation and uncertainty evaluation may not cover the real plant scale, the up-scaling capabilities of a best-estimate safety analysis code must be evaluated. The objective of this work is thus first of all, to refine the model parameters based on the Bayes’ theorem and subsequently estimate the uncertainties on parameters/responses during reflood phase. To illustrate this, reflood experiment data were collected and utilized to complete model calibration for the thermal–hydraulic parameters. The second goal of this study is to suggest optimum parameter distributions for the simulation of the multiple reflood tests performed at different facilities with different scales and dimensions. Since existing experimental data and physical models/correlations were produced from several tests performed at the small scale with limited initial and boundary conditions, scaling considerations must be addressed when simulating larger scale tests for the uncertainty analysis. Blind calculations were carried out to observe whether the a posteriori parameter samples obtained via the model calibration against a basis test, i.e., a small scale test, which was performed by compensating the scaling distortions properly simulate scaled up tests. Simulations were performed using Safety and Performance Analysis Code (SPACE) developed by multiple research institutes in Republic of Korea to predict the thermal hydraulic system behaviors of nuclear power plants. [ABSTRACT FROM AUTHOR]

This paper introduces the Poisson distribution, power-law distribution, and logarithmic-normal distribution as the prior distributions to construct Bayes statistical probability inference model for the simulation of the monthly crude oil price change point trends. Based on basic statistical cognition and product partition model (PPM), the historical change points are defined, identified, and analyzed. The PPM-KM integration model is established by combining PPM model and K-means method to measure, cluster, and identify the posterior probability of change points. The appearance probabilities of change points under different scenarios are calculated and compared for single recursive probabilistic predictions. The results showed there were 37 significant change points during 1986–2015. In different time points, unbalance of market supply-demand structure, sudden geopolitical event, the US dollar index, and global economic development situation have become the main reasons for oil price catastrophes. The monthly crude oil price change point complied with the power-law distribution hypothesis. It provides a new analytical perspective and is beneficial to governments, enterprises and investors to understand the market trends, avoid investment risks and make the right investment decisions. [ABSTRACT FROM AUTHOR]

Distributed drive electric vehicle(DDEV) has been widely researched recently, its longitudinal stability is a very important research topic. Conventional wheel slip ratio control strategies are usually designed for one special operating mode and the optimal performance cannot be obtained as DDEV works under various operating modes. In this paper, a novel model predictive controller-based multi-model control system (MPC-MMCS) is proposed to solve the longitudinal stability problem of DDEV. Firstly, the operation state of DDEV is summarized as three kinds of typical operating modes. A submodel set is established to accurately represent the state value of the corresponding operating mode. Secondly, the matching degree between the state of actual DDEV and each submodel is analyzed. The matching degree is expressed as the weight coefficient and calculated by a modified recursive Bayes theorem. Thirdly, a nonlinear MPC is designed to achieve the optimal wheel slip ratio for each submodel. The optimal design of MPC is realized by parallel chaos optimization algorithm(PCOA)with computational accuracy and efficiency. Finally, the control output of MPC-MMCS is computed by the weighted output of each MPC to achieve smooth switching between operating modes. The proposed MPC-MMCS is evaluated on eight degrees of freedom(8DOF)DDEV model simulation platform and simulation results of different condition show the benefits of the proposed control system. [ABSTRACT FROM AUTHOR]

The lithium-ion battery has become the main power source of many electronic devices, it is necessary to know its state-of-health and remaining useful life to ensure the reliability of electronic device. In this paper, a novel hybrid method with the thought of error-correction is proposed to predict the remaining useful life of lithium-ion battery, which fuses the algorithms of unscented Kalman filter, complete ensemble empirical mode decomposition (CEEMD) and relevance vector machine. Firstly, the unscented Kalman filter algorithm is adopted to obtain a prognostic result based on an estimated model and produce a raw error series. Secondly, a new error series is constructed by analyzing the decomposition results of the raw error series obtained by CEEMD method. Finally, the new error series is utilized by relevance vector machine regression model to predict the prognostic error which is adopted to correct the prognostic result obtained by unscented Kalman filter. Remaining useful life prediction experiments for batteries with different rated capacities and discharging currents are performed to show the high reliability of the proposed hybrid method. [ABSTRACT FROM AUTHOR]

For wind-sensitive structures, such as long-span bridges, high-rise buildings, transmission towers, etc., the prediction of wind speed and its statistical distribution are vital steps in the design and operation stages. Specifically, wind speed prediction is directly related to the value of wind load in the next occurrence; the statistical distribution of wind speed has regular characteristics, which can represent the random characteristics of wind field. In this paper, a probabilistic prediction model of wind speed based on Bayes' theorem is proposed and verified based on structural health monitoring (SHM) data. Firstly, the Gaussian process is derived and used as an a priori function in Bayes' theorem. In addition, the influence of six covariance functions on the prediction performance are discussed, that is, squared exponential (SE), Matern-3/2 (MA-3/2), Matern-5/2 (MA-5/2), automatic relevance determination SE (ARDSE), ARDMA-3/2, and ARDMA-5/2. Secondly, the correlation between the next wind speed and the previous wind speed is discussed by using the moving window method. Finally, the parameters in the three wind speed probability distribution functions (PDF), that is, Gumbel distribution, Weibull distribution, Rayleigh distribution, are updated in real time by increasing the SHM data based on Bayes' theorem. [ABSTRACT FROM AUTHOR]

In the Load and Resistance Factor Design (LRFD) of piles, several design codes recommend higher resistance factors if load tests are conducted. However, no information is provided on how these resistance factors are determined. In this paper, a probabilistic approach based on Bayes' theorem and the First Order Reliability Method (FORM) is proposed to calibrate resistance factors for different numbers of load tests and the corresponding test results. In addition, within-site variability, design methods, types of piles, and ground conditions can also be considered. The proposed approach applied to H-piles under axial load tests shows consistent results with current design codes. Results show that resistance factors are significantly increased even if only one positive test is observed among all the tests. For low variability sites, the differences of resistance factors between various design methods are significantly reduced if one or more tests are positive, while for high variability sites, the differences of resistance factors are only slightly decreased, indicating that design methods should be considered in the latter case. Most of the increase in resistance factors is achieved with a small number of tests. For β-Method used in clay sites, 80% of the increase in resistance factors is achieved with two, four and five consecutive positive tests are observed for low, medium and high variability sites, respectively. [ABSTRACT FROM AUTHOR]

The kNN algorithm remains a popular choice for pattern classification till date due to its non-parametric nature, easy implementation and the fact that its classification error is bounded by twice the Bayes error. In this paper, we show that the performance of the kNN classifier improves significantly from the use of (training) class-wise group-statistics based two criteria during pairwise comparison of features in a given dataset. Granger causality is employed to assign preferences to each criteria. Analytic Hierarchy Process (AHP) is applied to obtain weights for different features from the two criteria and their preferences. Finally, these weights are used to build a weighted distance function for the kNN classification. Comprehensive experimentation on fifteen benchmark datasets of the UCI Machine Learning Repository clearly reveals the supremacy of the proposed Granger causality driven AHP induced kNN algorithm over the kNN method with many different distance metrics, and, with various feature selection strategies. In addition, the proposed method is also shown to perform well on high-dimensional face and hand-writing recognition datasets. [ABSTRACT FROM AUTHOR]

Discovering user similarities from social media can establish the basis for user targeting, product recommendation, user relationship evolution and understanding. User similarities not only depend on the topological structure but also the dependence degrees between users. In this paper, we adopt Bayesian network (BN), an important and popular probabilistic graphical model, as the underling framework and propose a data-intensive approach for discovering user similarities. First, upon the massive social behavioral interactions, we give the method for measuring direct similarities between users and the MapReduce-based algorithm for constructing a BN to describe these similarities, called user Bayesian network and abbreviated as UBN. We also give the idea for storing large-scale UBNs in a distributed file system. Then, to measure indirect similarities between users, we give the method for measuring the closeness of user connections in terms of the properties of UBN's graphical structure. Further, we give the MapReduce-based algorithm for measuring the dependence degrees by means of UBN's probabilistic inferences. By combining the above two perspectives of measures, the indirect similarity degree between users can be achieved, while guaranteeing the applicability theoretically. Finally, we give experimental results and show the efficiency and effectiveness of our method. [ABSTRACT FROM AUTHOR]

In this paper, a new approach is proposed for coalbed methane (CBM) reservoir brittleness index ( BI ) calculations. The BI , as a guide for fracture area selection, is calculated by dynamic elastic parameters (dynamic Young's modulus E d and dynamic Poisson's ratio υ d ) obtained from an amplitude versus offset (AVO) simultaneous inversion. Among the three different classes of CBM reservoirs distinguished on the basis of brittleness in the theoretical part of this study, class I reservoirs with high BI values are identified as preferential target areas for fracturing. Therefore, we derive the AVO approximation equation expressed by E d and υ d first. This allows the direct inversion of the dynamic elastic parameters through the pre-stack AVO simultaneous inversion, which is based on Bayes' theorem. Thereafter, a test model with Gaussian white noise and a through-well seismic profile inversion is used to demonstrate the high reliability of the inversion parameters. Accordingly, the BI of a CBM reservoir section from the Qinshui Basin is calculated using the proposed method and a class I reservoir section detected through brittleness evaluation. From the outcome of this study, we believe the adoption of this new approach could act as a guide and reference for BI calculations and evaluations of CBM reservoirs. [ABSTRACT FROM AUTHOR]

• An adaptive Wiener process model with closed skew-normal distributed degradation drift is proposed. • EM algorithm with a novel Bayesian smoother based on CSN distribution is developed for model parameters estimation. • The analytical solution for the probability distribution function of RUL is derived. • The prognostic performance of the proposed model is testified by application to a real bearing dataset. Remaining useful life (RUL) prediction of rolling bearings is crucial to equipment operation and maintenance. The data-driven Wiener-based methods have aroused widespread concern. The commonly adopted distribution to represent inter-product variation in the Wiener degradation model is the normal distribution. However, this assumption is not appropriate for some practical applications. Aiming at this problem, an adaptive nonlinear Wiener process model with degradation drift satisfying the closed skew-normal (CSN) distribution to model the degradation characteristic exhibiting non-normality and asymmetry in inter-product variation is proposed in this paper. An online recursive algorithm based on Bayes' theorem is derived to update the hidden drift distribution. A Bayesian smoother based on CSN distribution is developed, which is employed to estimate model parameters involving the expectation–maximization algorithm. Furthermore, an analytical expression of the RUL distribution is derived. Finally, a numerical case and a practical bearing dataset are provided to illustrate the proposed method. [ABSTRACT FROM AUTHOR]

This overview paper describes the particle methods developed for the implementation of the class of Bayes filters formulated using the random finite set formalism. It is primarily intended for the readership already familiar with the particle methods in the context of the standard Bayes filter. The focus in on the Bernoulli particle filter, the probability hypothesis density (PHD) particle filter and the generalised labelled multi-Bernoulli (GLMB) particle filter. The performance of the described filters is demonstrated in the context of bearings-only target tracking application. [ABSTRACT FROM AUTHOR]

In this paper we deal with Bayes, E-Bayes and robust Bayes prediction under precautionary loss functions. It is well-known that in the Bayesian framework, the Bayes rule is obtained by considering a specific prior distribution over the parameter of interest but in practice, the use of a specified prior with specific hyperparameters is critical. Specially, when a problem in the Bayesian framework is behaved by two or more statisticians, they might agree on a specific prior but not on the hyperparameter choices. To deal with such an uncertainty issue, E-Bayes and robust Bayes approaches may be called, in which some classes of priors are considered and some optimal rules are derived. In this regard, we extend the idea of E-Bayes estimation to E-Bayes prediction and as a useful alternative approach, we deal with robust Bayes prediction. We also apply these approaches to the type-II censoring scheme. We conduct a simulation study and compare performance of the Bayes, E-Bayes and robust Bayes approaches. Finally, the proposed predictors are applied to a real data analysis and reporting some existing prediction methods in the literature, we illustrate practical utility of the prediction procedures. [ABSTRACT FROM AUTHOR]

In this paper, we examine Bayes factor consistency in the context of Bayesian variable selection for normal linear regression models. We take a hierarchical Bayesian approach using a hyper- g prior (Liang et al. (2008) JASA). There are two regimes for computing Bayes factors, which differ in the choice of the base model. For both these regimes, we study conditions under which Bayes factors are consistent when the number of all the potential regressors grows with sample size n . This situation is not fully understood in the current literature, but has gained increasing importance recently. In the present case, Bayes factors are not analytically tractable and are calculated via Laplace approximation. A rigorous justification for these high-dimensional Laplace approximations is also provided. [ABSTRACT FROM AUTHOR]

In many areas of science, models are used to describe attributes of complex systems. These models are generally themselves highly complex functions of their inputs, and can be computationally expensive to evaluate. Often, these models have parameters which must be estimated using data from the real system. In this paper, we address the problem of using prior information supplied by the model, in conjunction with prior beliefs about its parameters, to design the collection of data such that it is optimal for decisions which must be made using posterior beliefs about the model parameters. Optimal design calculations do not generally have a closed form solution, so we propose a Bayes linear analysis to find an approximately optimal design. We motivate the approach by considering optimal specification of measurement locations for remote sensing of airborne species. [ABSTRACT FROM AUTHOR]

Predicted performance of a complex geotechnical system is subject to errors due to the uncertainties associated with both the prediction model and the system parameters. This paper aims to develop a multi-step updating method to reduce the uncertainties of the prediction model and system parameters using incremental time-series monitoring data, and to enhance the prediction of the future performance of the complex geotechnical system. The multi-step updating method considers inherent uncertainty of the system, model uncertainty and measurement uncertainty. The prediction is updated and improved step by step with new monitoring information using Bayes’ theorem. Two realistic geotechnical cases including a basement excavation and a multi-stage excavation of a high rock slope are presented for illustration. The multi-step updating method integrates the theoretical computational model with observational evidence. With more monitoring information being incorporated, the prediction becomes closer to the actual performance of the system, and the distributions of the system parameters become closer to the reality. [ABSTRACT FROM AUTHOR]

Identification and assessment of audit risks, which may be encountered by a company, is a very important step in the audit procedure. The risk is actually a threat to society, something that could happen, the threat resulting from a specified event, so something has to happen for the risk to become “alive”. The key to success is knowledge, but the key to success in the global context is the ability to adapt your knowledge to the trend. If, until now, methods for risk identification and assessment in financial audit used by auditors are classic, they need to upgrade and level up to the next step. The present paper presents three methods for risk assessment in financial auditing: ARM, Bayes’ theorem and belief function, managing to conclude that for adapting to the technology that it is in constantly change, computer-assisted audit techniques are becoming slowly –slowly a necessary evolution. [ABSTRACT FROM AUTHOR]

It has become increasingly important to understand the asymptotic behavior of the Bayes factor for model selection in general statistical models. In this paper, we discuss recent results on Bayes factor consistency in semiparametric regression problems where observations are independent but not identically distributed. Specifically, we deal with the model selection problem in the context of partial linear models in which the regression function is assumed to be the additive form of the parametric component and the nonparametric component using Gaussian process priors, and Bayes factor consistency is investigated for choosing between the parametric model and the semiparametric alternative. [ABSTRACT FROM AUTHOR]

The marginal distribution Bayes (MDB) filter is an efficient approach for tracking an unknown and time-varying number of targets in the presence of clutter, noise, data association uncertainty, and detection uncertainty. This filter propagates the marginal distributions and existence probabilities of each target in the filter recursion, and it admits a closed-form solution for a linear Gaussian multi-target model. However, this closed-form solution is not general enough to accommodate nonlinear multi-target models. In this paper, we propose two implementations of the MDB filter to accommodate nonlinear multi-target models. The first is the first-order Taylor approximation MDB (FTA-MDB) filter which is based on the linearization technique of nonlinear function, and the second is the unscented transform MDB (UT-MDB) filter which is based on the unscented transform technique. Simulation results demonstrate that the proposed implementations are better on multiple targets tracking than the UK-PHD filter. [ABSTRACT FROM AUTHOR]

In our experience, it seems to become more and more common for mandating authorities or parties to ask forensic signature examiners to quantify the degree of certainty of their conclusion regarding a signature analysis. This paper reports the likelihood ratio approach followed by examiners to answer such a question, in a case where the Court asked whether a questioned signature was written, or not, by Mr Jones. The Court also required an assessment of the error margin of the signature analysis. This question was answered using Bayes' theorem (i.e., a full Bayesian approach) and this paper seeks to show that such an approach can be used despite the popular belief that Bayes' theorem is beyond what courts may accept. Using a practical example, we present advantages of the approach we have chosen to assess our results and show that a logical approach for evidence evaluation can be followed even in a forensic discipline where no tabulated data are available. This example also illustrates a practical way of addressing the error margin question, which helps the Court understand what can be the risk of being wrong in this particular case (and not in cases in general). We further present the way these results were communicated to the fact finders in the case at hand and provide guidance as how forensic observations can logically be combined with the other elements of the case. [ABSTRACT FROM AUTHOR]

• A review of the machine learning-based structural reliability analysis methods is presented. • Artificial neural networks-based structural reliability analysis methods are reviewed. • Support vector machines-based structural reliability analysis methods are reviewed. • The application of Bayes' theorem for structural reliability analysis is reviewed. • The application of Kriging estimation with an active learning perspective for structural reliability analysis is reviewed. Structural Reliability analysis (SRA) is one of the prominent fields in civil and mechanical engineering. However, an accurate SRA in most cases deals with complex and costly numerical problems. Machine learning-based (ML) techniques have been introduced to the SRA problems to deal with this huge computational cost and increase accuracy. This paper presents a review of the development and use of ML models in SRA. The review includes the most common types of ML methods used in SRA. More specifically, the application of artificial neural networks (ANN), support vector machines (SVM), Bayesian methods and Kriging estimation with active learning perspective in SRA are explained, and a state-of-the-art review of the prominent literature in these fields is presented. Aiming towards a fast and accurate SRA, the ML techniques adopted for the approximation of the limit state function with Monte Carlo simulation (MCS), first/second-order reliability methods (FORM/SORM) or MCS with importance sampling well as the methods for efficiently computing the probabilities of rare events in complex structural systems. In this regard, the focus of the current manuscript is on the different models' structures and diverse applications of each ML method in different aspects of SRA. Moreover, imperative considerations on the management of samples in the Monte Carlo simulation for SRA purposes and the treatment of the SRA problem as pattern recognition or classification task are provided. This review helps the researchers in civil and mechanical engineering, especially those who are focused on reliability and structural analysis or dealing with product assurance problems. [ABSTRACT FROM AUTHOR]