Adaptive Sampling Approach Research Articles

An Adaptive Sampling Approach for Evaluating Robot Self-Righting Capabilities

To properly evaluate the ability of robots to operate autonomously in the real world, it is necessary to develop methods for quantifying their self-righting capabilities. Here, we improve upon a sampling-based framework for evaluating self-righting capabilities that was previously validated in two dimensions. To apply this framework to realistic robots in three dimensions, we require algorithms capable of scaling to high-dimensional configuration spaces. Therefore, we introduce a novel adaptive sampling approach that biases queries toward the transitional states of the system, thus, identifying the critical transitions of the system using substantially fewer samples. To demonstrate this improvement, we compare our approach to results that were generated via the previous framework and were validated on hardware platforms. Finally, we apply our technique to a high-fidelity three-dimensional model of a US Navy bomb-defusing robot, which was too complex for the previous framework to analyze.

Integrating synthetic minority oversampling and gradient boosting decision tree for bogie fault diagnosis in rail vehicles

Bogies are critical components of a rail vehicle, which are important for the safe operation of rail transit. In this study, the authors analyzed the real vibration data of the bogies of a railway vehicle obtained from a Chinese subway company under four different operating conditions. The authors selected 15 feature indexes – that ranged from time-domain, energy, and entropy – as well as their correlations. The adaptive synthetic sampling approach–gradient boosting decision tree (ADASYN–GBDT) method is proposed for the bogie fault diagnosis. A comparison between ADASYN–GBDT and the three commonly used classifiers (K-nearest neighbor, support vector machine, and Gaussian naïve Bayes), combined with random forest as the feature selection, was done under different test data sizes. A confusion matrix was used to evaluate those classifiers. In K-nearest neighbor, support vector machine, and Gaussian naïve Bayes, the optimal features should be selected first, while the proposed method of this study does not need to select the optimal features. K-nearest neighbor, support vector machine, and Gaussian naïve Bayes produced inaccurate results in multi-class identification. It can be seen that the lowest false detection rates of the proposed ADASYN–GBDT model are 92.95% and 87.81% when proportion of the test dataset is 0.4 and 0.9, respectively. In addition, the ADASYN–GBDT model has the ability to correctly identify a fault, which makes it more practical and suitable for use in railway operations. The entire process (training and testing) was finished in 2.4231 s and the detection procedure took 0.0027 s on average. The results show that the proposed ADASYN–GBDT method satisfied the requirements of real-time performance and accuracy for online fault detection. It might therefore aid in the fault detection of bogies.

Plotting the map projection graticule involving discontinuities based on combined sampling

This article presents new algorithm for interval plotting the projection graticule on the interval $\varOmega=\varOmega_{\varphi}\times\varOmega_{\lambda}$ based on the combined sampling technique. The proposed method synthesizes uniform and adaptive sampling approaches and treats discontinuities of the coordinate functions $F,G$. A full set of the projection constant values represented by the projection pole $K=[\varphi_{k},\lambda_{k}]$, two standard parallels $\varphi_{1},\varphi_{2}$ and the central meridian shift $\lambda_{0}^{\prime}$ are supported. In accordance with the discontinuity direction it utilizes a subdivision of the given latitude/longitude intervals $\varOmega_{\varphi}=[\underline{\varphi},\overline{\varphi}]$, $\varOmega_{\lambda}=[\underline{\lambda},\overline{\lambda}]$ to the set of disjoint subintervals $\varOmega_{k,\varphi}^{g},$$\varOmega_{k,\lambda}^{g}$ forming tiles without internal singularities, containing only "good" data; their parameters can be easily adjusted. Each graticule tile borders generated over $\varOmega_{k}^{g}=\varOmega_{k,\varphi}^{g}\times\varOmega_{k,\lambda}^{g}$ run along singularities. For combined sampling with the given threshold $\overline{\alpha}$ between adjacent segments of the polygonal approximation the recursive approach has been used; meridian/parallel offsets are $\Delta\varphi,\Delta\lambda$. Finally, several tests of the proposed algorithms are involved.

Efficient plotting the functions with discontinuities based on combined sampling

This article presents a new algorithm for interval plotting of the function y = f(x) based on combined sampling. The proposed method synthesizes the uniform and adaptive sampling approaches and provides a more compact and efficient function representation. During the combined sampling, the polygonal approximation with a given threshold α between the adjacent segments is constructed. The automated detection and treatment of the discontinuities based on the LR criterion are involved. Two implementations, the recursive-based and stack-based, are introduced. Finally, several tests of the proposed algorithms for the different functions involving the discontinuities and several map projection graticules are presented. The proposed method may be used for more efficient sampling the curves (map projection graticules, contour lines, or buffers) in geoinformatics.

Constrained optimization of black-box stochastic systems using a novel feasibility enhanced Kriging-based method

Stochastically constrained simulation optimization problems are challenging because the inherent noise terms to a black-box system lead to the need of considering the uncertainty at both the objective and the constraint function. To address this problem, we propose a Kriging-based optimization framework, which uses stochastic Kriging models to approximate the objective and the constraint functions and an adaptive sampling approach to sequentially search for the next point that is promising to have a better objective. The main contribution of this work is to incorporate a feasibility-enhanced term to the infill sampling criterion, which improves the Kriging-based algorithm's capabilities of returning a truly feasible near-optimal solution for stochastic systems. The efficacy of the Kriging-based algorithm is demonstrated with eight benchmark problems and a case study from the pharmaceutical manufacturing domain.

Cross entropy-based importance sampling using Gaussian densities revisited

The computation of the probability of a rare (failure) event is a common task in structural reliability analysis. In most applications, the numerical model defining the rare event is nonlinear and the resulting failure domain often multimodal. One strategy for estimating the probability of failure in this context is the importance sampling method. The efficiency of importance sampling depends on the choice of the importance sampling density. A near-optimal sampling density can be found through application of the cross entropy method. The cross entropy method is an adaptive sampling approach that determines the sampling density through minimizing the Kullback-Leibler divergence between the theoretically optimal importance sampling density and a chosen parametric family of distributions. In this paper, we investigate the suitability of the multivariate normal distribution and the Gaussian mixture model as importance sampling densities within the cross entropy method. Moreover, we compare the performance of the cross entropy method to sequential importance sampling, another recently proposed adaptive sampling approach, which uses the Gaussian mixture distribution as a proposal distribution within a Markov Chain Monte Carlo algorithm. For the parameter updating of the Gaussian mixture within the cross entropy method, we propose a modified version of the expectation-maximization algorithm that works with weighted samples. To estimate the number of distributions in the mixture, the density-based spatial clustering of applications with noise (DBSCAN) algorithm is adapted to the use of weighted samples. We compare the performance of the different methods in several examples, including component reliability problems, system reliability problems and reliability in varying dimensions. The results show that the cross entropy method using a single Gaussian outperforms the cross entropy method using Gaussian mixture and that both distribution types are not suitable for high dimensional reliability problems.

Adaptive stratified importance sampling: hybridization of extrapolation and importance sampling Monte Carlo methods for estimation of wind turbine extreme loads

Abstract. Wind turbine extreme load estimation is especially difficult because turbulent inflow drives nonlinear turbine physics and control strategies; thus there can be huge differences in turbine response to essentially equivalent environmental conditions. The two main current approaches, extrapolation and Monte Carlo sampling, are both unsatisfying: extrapolation-based methods are dangerous because by definition they make predictions outside the range of available data, but Monte Carlo methods converge too slowly to routinely reach the desired 50-year return period estimates. Thus a search for a better method is warranted. Here we introduce an adaptive stratified importance sampling approach that allows for treating the choice of environmental conditions at which to run simulations as a stochastic optimization problem that minimizes the variance of unbiased estimates of extreme loads. Furthermore, the framework, built on the traditional bin-based approach used in extrapolation methods, provides a close connection between sampling and extrapolation, and thus allows the solution of the stochastic optimization (i.e., the optimal distribution of simulations in different wind speed bins) to guide and recalibrate the extrapolation. Results show that indeed this is a promising approach, as the variance of both the Monte Carlo and extrapolation estimates are reduced quickly by the adaptive procedure. We conclude, however, that due to the extreme response variability in turbine loads to the same environmental conditions, our method and any similar method quickly reaches its fundamental limits, and that therefore our efforts going forward are best spent elucidating the underlying causes of the response variability.

Cruise missile head shape optimisation using an adaptive sampling surrogate model

ABSTRACTHigh-precision response of the surrogate model is desired in the process of optimisation. An excessive number of sampling points will increase the cost of the calculation. The appropriate number of sampling points cannot only guarantee the accuracy of the surrogate model but also save the calculation cost. The purpose of this research is to demonstrate the eventuality of using an adaptive surrogate model for optimisation problems. The adaptive surrogate model is built on an adaptive sampling approach and an extended radial basis function (ERBF). The adaptive sampling is an approach that new sampling points are placed in the blank area and all the sampling points are uniformly distributed in the design region using Multi-Island GA. The precision of the ERBF surrogate model is checked using standard error measure to determine whether the surrogate model should be updated or not. This adaptive surrogate model is used to optimise a cruise missile head shape. Aerodynamic and stealthy performance of the cruise missile head shape are considered in this research. Different global objective function and different weight factor are used to research the aerodynamic and stealthy performance in this optimisation process. The results show that the drag is reduced with a slender head shape and the radar-cross section (RCS) value is reduced with a short head shape.

Health stages diagnostics of underwater thruster using sound features with imbalanced dataset

The underwater thruster is considered one of the most critical components located on an unmanned underwater vehicle to maneuver in the water. However, it is recognized as a common source of the fault. This phenomenon is made worse when collected data for equipment health diagnostics are highly imbalanced. A new sampling method to tackle the problem of imbalanced data based on cosine similarity is proposed to improve the classification accuracy for thruster health diagnostics. The results show that it outperforms SMOTE (Synthetic Minority Oversampling Technique) and ADASYN (Adaptive Synthetic Sampling Approach for Imbalanced Learning). The proposed method was further validated using different imbalanced datasets with different imbalance ratio from KEEL and UCI machine learning repository (such as Pima Indians Diabetes, Ionosphere, Fertility Diagnostics, Mammographic Masses, Blood Transfusion Service Centre). The majority of the results from the datasets show that the proposed method produces the higher classification accuracy as well as g-means that suggests the potential approach for classification problem that has a highly imbalanced dataset.

Multiple player tracking in soccer videos: an adaptive multiscale sampling approach

Visual tracking is an essential technique in computer vision. Even though the notable improvement has been achieved during last few years, tracking multiple objects still remains as a challenging task. In this paper, a novel method for tracking multiple players in soccer videos, which include severe occlusions between players and nonlinear motions by their complex interactions, is introduced. Specifically, we first extract moving objects (i.e., players) by refining results of background subtraction via the edge information obtained from the frame differencing result. Then, we conduct multiscale sampling in foreground regions, which are spatially close to each tracked player, and subsequently computing the dissimilarity between sampled image blocks and each tracked player. Based on the best-matched case, the state of each tracked player (e.g., center position, color, etc.) is consistently updated using the online interpolation scheme. Experimental results in various soccer videos show the efficiency and robustness of our method compared to previous approaches introduced in the literature.

An adaptive sampling approach to reduce uncertainty in slope stability analysis

An adaptive sampling approach is proposed, which can sample spatially varying shear strength parameters efficiently to reduce uncertainty in the slope stability analysis. This approach employs a limit equilibrium model and stochastic conditional methodology to determine the likely sampling locations. Karhunen-Loeve expansion is used to conduct the conditional Monte Carlo simulation. A first-order analysis is also proposed to ease the computational burden associated with Monte Carlo simulation. These approaches are then tested using borehole data from a field site. Results indicate that the proposed adaptive sampling approach is an effective and efficient sampling scheme for reducing uncertainty in slope stability analysis.

Driving behavior recognition using EEG data from a simulated car-following experiment

Driving behavior recognition is the foundation of driver assistance systems, with potential applications in automated driving systems. Most prevailing studies have used subjective questionnaire data and objective driving data to classify driving behaviors, while few studies have used physiological signals such as electroencephalography (EEG) to gather data. To bridge this gap, this paper proposes a two-layer learning method for driving behavior recognition using EEG data. A simulated car-following driving experiment was designed and conducted to simultaneously collect data on the driving behaviors and EEG data of drivers. The proposed learning method consists of two layers. In Layer I, two-dimensional driving behavior features representing driving style and stability were selected and extracted from raw driving behavior data using K-means and support vector machine recursive feature elimination. Five groups of driving behaviors were classified based on these two-dimensional driving behavior features. In Layer II, the classification results from Layer I were utilized as inputs to generate a k-Nearest-Neighbor classifier identifying driving behavior groups using EEG data. Using independent component analysis, a fast Fourier transformation, and linear discriminant analysis sequentially, the raw EEG signals were processed to extract two core EEG features. Classifier performance was enhanced using the adaptive synthetic sampling approach. A leave-one-subject-out cross validation was conducted. The results showed that the average classification accuracy for all tested traffic states was 69.5% and the highest accuracy reached 83.5%, suggesting a significant correlation between EEG patterns and car-following behavior.

Deep convolutional neural network with transfer learning for rectum toxicity prediction in cervical cancer radiotherapy: a feasibility study

Better understanding of the dose-toxicity relationship is critical for safe dose escalation to improve local control in late-stage cervical cancer radiotherapy. In this study, we introduced a convolutional neural network (CNN) model to analyze rectum dose distribution and predict rectum toxicity. Forty-two cervical cancer patients treated with combined external beam radiotherapy (EBRT) and brachytherapy (BT) were retrospectively collected, including twelve toxicity patients and thirty non-toxicity patients. We adopted a transfer learning strategy to overcome the limited patient data issue. A 16-layers CNN developed by the visual geometry group (VGG-16) of the University of Oxford was pre-trained on a large-scale natural image database, ImageNet, and fine-tuned with patient rectum surface dose maps (RSDMs), which were accumulated EBRT + BT doses on the unfolded rectum surface. We used the adaptive synthetic sampling approach and the data augmentation method to address the two challenges, data imbalance and data scarcity. The gradient-weighted class activation maps (Grad-CAM) were also generated to highlight the discriminative regions on the RSDM along with the prediction model. We compare different CNN coefficients fine-tuning strategies, and compare the predictive performance using the traditional dose volume parameters, e.g. D0.1/1/2cc, and the texture features extracted from the RSDM. Satisfactory prediction performance was achieved with the proposed scheme, and we found that the mean Grad-CAM over the toxicity patient group has geometric consistence of distribution with the statistical analysis result, which indicates possible rectum toxicity location. The evaluation results have demonstrated the feasibility of building a CNN-based rectum dose-toxicity prediction model with transfer learning for cervical cancer radiotherapy.

A Novel Surrogate-Based Optimization Method for Black-Box Simulation with Heteroscedastic Noise

Simulation optimization (SO) problems can be difficult to solve because of the lack of knowledge of the algebraic model equations and the unknown structure of the noise inherent to the simulation. It is important to investigate approaches capable of handling noise in order to achieve optimal solution with efficiency. In recent years, surrogate-based methods for SO problems have gained increasing attention from different research communities. In this work, we adapted a one-stage adaptive sampling approach to a Kriging-based optimization framework for simulations with heteroscedastic noise. We compared its performance with another Kriging-based approach using expected improvement as the infill criterion. On the basis of the results of several test problems, each with various noise scenarios, we discussed the benefits and limitations of both algorithms. Finally, we show the application of both algorithms to finding the optimal operation conditions of a continuous pharmaceutical manufacturing simulation model.

GIR-based ensemble sampling approaches for imbalanced learning

This paper presents two adaptive ensemble sampling approaches for imbalanced learning: one is the undersampling-based approach, and the other one is the oversampling-based approach, with the objectives of bias reduction and adaptive learning. Both of these two approaches are based on a novel class imbalance metric, termed generalized imbalance ratio (GIR), instead of the conventional sample size ratio. Specifically, these two sampling-based approaches adaptively split the imbalanced learning problem into multiple balanced learning subproblems in a probabilistic way, which forces the classifiers trained in the subproblems focus on those difficult to learn samples. In each subproblem, several weak classifiers are trained in a boosting manner. A final stronger classifier is further built by combining all these weak classifiers in a bagging manner. Extensive experiments are conducted on real-life UCI imbalanced data sets to evaluate the performance of the proposed methods. The superior performance demonstrates the effectiveness of the proposed methods and indicates wide potential applications in data mining.

An adaptive sampling approach for Kriging metamodeling by maximizing expected prediction error

As a well-known approximation method, Kriging is widely used in process engineering design and optimization for saving computational budget. The Kriging model for a target function is fitted to a set of sample points, the responses of which are expensive to obtain in practice and the sample distribution of which has a great impact on the model prediction quality. Therefore, a main task in adaptive sampling for Kriging metamodeling is to gather informative points in order to build an accurate model with as few points as possible. To this end, we propose an adaptive sampling approach under the bias-variance decomposition framework. This novel sampling approach sequentially selects new points by maximizing an expected prediction error criterion that considers both the bias and variance information. Particularly, it presents an adaptive balance strategy to dynamically balance the local exploitation and global exploration via the error information from the previous iteration. Four benchmark cases and four engineering cases from low to high dimensions are used to assess the performance of the proposed approach. Numerical results reveal that this adaptive sampling approach is very promising for constructing accurate Kriging models for problems with diverse characteristics.

Adaptive Data Collection protocol for Extending Lifetime of Periodic Sensor Networks

In Periodic Sensor Network (PSN), the periodic energy efficient collection of a large amount of sensed data by the sensor nodes is considered as one of the main challenges in this type of network. An adaptive sampling approach to periodic data collection is needed for energy optimization due to the limited lifetime nature of the sensor batteries. In this paper, we propose a protocol, called an Adaptive Data Collection protocol (ADaC), which collects periodically sensor readings and prolong the lifetime of a Periodic Sensor Network (PSN). The lifetime of ADaC protocol is divided into cycles. Each cycle is composed of two stages.First, data collection. Second, sampling rate adaptation based on the similarity between periods of one cycle using Euclidean distance measure to adapt its rate of sampling according to the dynamic modification of the monitored environment. ADaC allows each sensor to adapt its sampling rate in accordance with the monitored environment conditions. We conduct extensive simulation experiments on real sensor data by applying OMNeT++ network simulator to explain the effectiveness of the ADaC protocol in comparison with a method without using adaptive sampling.

An Adaptive Metamodel-Based Subset Importance Sampling approach for the assessment of the functional failure probability of a thermal-hydraulic passive system

An Adaptive Metamodel-Based Subset Importance Sampling (AM-SIS) approach, previously developed by the authors, is here employed to assess the (small) functional failure probability of a thermal-hydraulic (T-H) nuclear passive safety system. The approach relies on an iterative Importance Sampling (IS) scheme that efficiently couples the powerful characteristics of Subset Simulation (SS) and fast-running Artificial Neural Networks (ANNs). In particular, SS and ANNs are intelligently employed to build and progressively refine a fully nonparametric estimator of the ideal, zero-variance Importance Sampling Density (ISD), in order to: (i) increase the robustness of the failure probability estimates and (ii) decrease the number of expensive T-H simulations needed (together with the related computational burden). The performance of the approach is thoroughly compared to that of other efficient Monte Carlo (MC) techniques on a case study involving an advanced nuclear reactor cooled by helium.

Diagnosis of retinal health in digital fundus images using continuous wavelet transform (CWT) and entropies

Vision is paramount to humans to lead an active personal and professional life. The prevalence of ocular diseases is rising, and diseases such as glaucoma, Diabetic Retinopathy (DR) and Age-related Macular Degeneration (AMD) are the leading causes of blindness in developed countries. Identifying these diseases in mass screening programmes is time-consuming, labor-intensive and the diagnosis can be subjective. The use of an automated computer aided diagnosis system will reduce the time taken for analysis and will also reduce the inter-observer subjective variabilities in image interpretation. In this work, we propose one such system for the automatic classification of normal from abnormal (DR, AMD, glaucoma) images. We had a total of 404 normal and 1082 abnormal fundus images in our database. As the first step, 2D-Continuous Wavelet Transform (CWT) decomposition on the fundus images of two classes was performed. Subsequently, energy features and various entropies namely Yager, Renyi, Kapoor, Shannon, and Fuzzy were extracted from the decomposed images. Then, adaptive synthetic sampling approach was applied to balance the normal and abnormal datasets. Next, the extracted features were ranked according to the significances using Particle Swarm Optimization (PSO). Thereupon, the ranked and selected features were used to train the random forest classifier using stratified 10-fold cross validation. Overall, the proposed system presented a performance rate of 92.48%, and a sensitivity and specificity of 89.37% and 95.58% respectively using 15 features. This novel system shows promise in detecting abnormal fundus images, and hence, could be a valuable adjunct eye health screening tool that could be employed in polyclinics, and thereby reduce the workload of specialists at hospitals.

Automated screening tool for dry and wet age-related macular degeneration (ARMD) using pyramid of histogram of oriented gradients (PHOG) and nonlinear features

Aging is the prime cause of age-related macular degeneration (ARMD). There are primarily two types of ARMD: (i) dry, and (ii) wet. The dry ARMD is the common form of ARMD. It typically starts with the formation of small pale yellowish deposits called drusen under the retina, causing atrophy at the macula, which is known as age-related macular degeneration. This, in turn, affects the central vision of a person. The wet ARMD is caused due to the abnormal growth of blood vessels under the retina. These vessels are known as the choroidal neovascular membrane break through the retina and bleed. They finally lead to either scarring at the macula or atrophic changes leading to severe visual impairment. Hence, the wet ARMD progresses faster and leads to an irreversible loss of sight. Therefore, it is advisable to go for routine eye screening, especially for elderly subjects. Although the ARMD cannot be fully cured, an accurate early detection can impede the progression of vision loss. However, manual diagnosis of ARMD is difficult and subjective. Thus, a computer-aided diagnosis (CAD) system can be utilized to automatically screen the eyes and give an accurate diagnosis of the type of ARMD. In this study, we propose a novel technique to identify normal, dry, and wet ARMD. A total of 945 fundus images (404 normal, 517 dry ARMD, and 24 wet ARMD) are used in this proposed framework. The Pyramid of Histograms of Orientation Gradients (PHOG) technique is implemented in this work to capture the subtle changes in the pixels of fundus images. Various nonlinear features are extracted from the PHOG descriptor. To balance the number of images in three classes, an adaptive synthetic sampling (ADASYN) approach is used. Two feature selection techniques namely ant colony optimization genetic algorithm (ACO-GA) and particle swarm optimization (PSO) are used to select the best performing method. The selected features are subjected to analysis of variance (ANOVA) to determine highly significant features for classification. The proposed system has achieved a maximum accuracy of 85.1%, sensitivity of 87.2%, and specificity of 80% using nine features selected by PSO feature selection method with support vector machine (SVM) classifier. It has obtained a maximum accuracy of 83.3%, sensitivity of 82.6%, and specificity of 84.8% using ten features selected by ACO-GA feature selection method with SVM classifier. The proposed CAD system yielded promising performance and thus, can be used as a practical adjunct ARMD screening tool in the clinical setting.