Efficient Sampling Technique Research Articles

Likelihood‐Based Inference and Prediction in Spatio‐Temporal Panel Count Models for Urban Crimes

SummaryWe develop a panel count model with a latent spatio‐temporal heterogeneous state process for monthly severe crimes at the census‐tract level in Pittsburgh, Pennsylvania. Our dataset combines Uniform Crime Reporting data with socio‐economic data. The likelihood is estimated by efficient importance sampling techniques for high‐dimensional spatial models. Estimation results confirm the broken‐windows hypothesis whereby less severe crimes are leading indicators for severe crimes. In addition to ML parameter estimates, we compute several other statistics of interest for law enforcement such as spatio‐temporal elasticities of severe crimes with respect to less severe crimes, out‐of‐sample forecasts, predictive distributions and validation test statistics. Copyright © 2016 John Wiley & Sons, Ltd.

Role of Statistical and Machine Learning Methods in Special Protection Scheme Logic Design and Failure Assessment

With tremendous growth in variable renewable generation and uncertainties in power system operations, there has been lot of installations of special protection schemes in many parts of the world. A special protection scheme, also termed a remedial action scheme, is the high-end last line of defense mechanism to relieve the system from catastrophic consequences under highly stressed conditions. This article delineates the special protection scheme logic design process and, with the help of a Nordic grid case study, proposes a new paradigm of special protection scheme failure assessment and logic design process, namely the “system-view” framework. In this context, the article also demonstrates the application of decision tree-based methods to derive special protection scheme operational logic and provides a pathway to use efficient statistical sampling techniques for implementing the system-view special protection scheme failure assessment method, thereby advancing the logic redesign process.

Performance modeling of CMOS inverters using support vector machines (SVM) and adaptive sampling

Integrated circuit designs are verified through the use of circuit simulators before being reproduced in real silicon. In order for any circuit simulation tool to accurately predict the performance of a CMOS design, it should generate models to predict the transistor’s electrical characteristics. The circuit simulation tools have access to massive amounts of data that are not only dynamic but generated at high speed in real time, hence making fast simulation a bottleneck in integrated circuit design. Using all the available data is prohibitive due to memory and time constraints. Accurate and fast sampling has been shown to enhance processing of large datasets without knowing all of the data. However, it is difficult to know in advance what size of the sample to choose in order to guarantee good performance. Thus, determining the smallest sufficient dataset size that obtains the same accurate model as the entire available dataset remains an important research question. This paper focuses on adaptively determining how many instances to present to the simulation tool for creating accurate models. We use Support Vector Machines (SVMs) with Chernoff inequality to come up with an efficient adaptive sampling technique, for scaling down the data. We then empirically show that the adaptive approach is faster and produces accurate models for circuit simulators as compared to other techniques such as progressive sampling and Artificial Neural Networks.

Evaluation of invertebrate diversity on a broad bean crop (Vicia fabaL. var. major) in Tizi-Ouzou area (Algeria)

AbstractThis study deals with the invertebrate biodiversity restricted to a broad bean crop, Vicia faba L. var. major, in Tizi-Ouzou (Algeria). The different sampling methods used (Barber pot, coloured traps, and sweep net) allowed drawing up a list of invertebrates collected within the study area. The inventory showed the presence of 54 species belonging to four classes, namely Gastropoda, Crustacea, Arachnida, and Insecta. Insecta was the best represented class with 11 different orders, the most important of which were Hymenoptera and Homoptera. In terms of total invertebrate species richness, Barber pot trapping was the most efficient sampling technique which collected 248 individuals belonging to 47 species; Messor barbarus was the most noted species with 24 individuals (F.c.% = 9.67%).The captured invertebrates were more frequent in the yellow plates with 269 individuals belonging to 40 species; Aphis fabae was the most noted with 48 individuals (F.c.% = 17.84%). The sweep net method allowed collecti...

Inverse estimation of thermal properties using Bayesian inference and three different sampling techniques

In this article, three different thermal properties such as thermal conductivity, heat transfer coefficient and emissivity are retrieved simultaneously using Bayesian inverse framework. Metropolis–Hasting Markov Chain Monte-Carlo (MH-MCMC) sampling is more commonly used in the literature to sample through posterior probability distribution function (PPDF) to find the expectations such as mean, standard deviation. However, when the posterior is multi-model/correlated, sometimes MH-MCMC struck with one mode and fails to sample through other modes which have significant probability. Nevertheless, efficient sampling techniques are being developed during the last decade to overcome this problem. Therefore, in the present work two population-based sampling techniques such as Parallel Tempering (PT) and Evolutionary Monte-Carlo (EMC) are used along with MH-MCMC to sample through correlated PPDF to retrieve the above three thermal properties. The estimation is carried out at three levels of measurement errors. The experimental data are obtained by adding normally distributed random noise with mean at zero and standard deviation ‘’ to the exact solution of the forward problem. The results show that when the measurement error is zero, i.e when , all three techniques perform well and estimates the parameter reasonably good. However, at and 0.5, MCMC estimates the parameter poorly, whereas both PT and EMC performs equally and estimates the parameter within a maximum of 9% deviation from the exact value.

Use of magnetic effervescent tablet-assisted ionic liquid dispersive liquid–liquid microextraction to extract fungicides from environmental waters with the aid of experimental design methodology

In this work, a novel effervescence-assisted microextraction technique was proposed for the detection of four fungicides. This method combines ionic liquid-based dispersive liquid–liquid microextraction with the magnetic retrieval of the extractant. A magnetic effervescent tablet composed of Fe3O4 magnetic nanoparticles, sodium carbonate, sodium dihydrogen phosphate and 1-hexyl-3-methylimidazolium bis(trifluoromethanesulfonimide) was used for extractant dispersion and retrieval. The main factors affecting the extraction efficiency were screened by a Plackett–Burman design and optimized by a central composite design. Under the optimum conditions, good linearity was obtained for all analytes in pure water model and real water samples. Just for the pure water, the recoveries were between 84.6% and 112.8%, the limits of detection were between 0.02 and 0.10 μg L−1 and the intra-day precision and inter-day precision both are lower than 4.9%. This optimized method was successfully applied in the analysis of four fungicides (azoxystrobin, triazolone, cyprodinil, trifloxystrobin) in environmental water samples and the recoveries ranged between 70.7% and 105%. The procedure promising to be a time-saving, environmentally friendly, and efficient field sampling technique.

A benchmark study on intelligent sampling techniques in Monte Carlo simulation

In recent years, new, intelligent and efficient sampling techniques for Monte Carlo simulation have been developed. However, when such new techniques are introduced, they are compared to one or two existing techniques, and their performance is evaluated over two or three problems. A literature survey shows that benchmark studies, comparing the performance of several techniques over several problems, are rarely found. This article presents a benchmark study, comparing Simple or Crude Monte Carlo with four modern sampling techniques: Importance Sampling Monte Carlo, Asymptotic Sampling, Enhanced Sampling and Subset Simulation; which are studied over six problems. Moreover, these techniques are combined with three schemes for generating the underlying samples: Simple Sampling, Latin Hypercube Sampling and Antithetic Variates Sampling. Hence, a total of fifteen sampling strategy combinations are explored herein. Due to space constrains, results are presented for only three of the six problems studied; conclusions, however, cover all problems studied. Results show that Importance Sampling using design points is extremely efficient for evaluating small failure probabilities; however, finding the design point can be an issue for some problems. Subset Simulation presented very good performance for all problems studied herein. Although similar, Enhanced Sampling performed better than Asymptotic Sampling for the problems considered: this is explained by the fact that in Enhanced Sampling the same set of samples is used for all support points; hence a larger number of support points can be employed without increasing the computational cost. Finally, the performance of all the above techniques was improved when combined with Latin Hypercube Sampling, in comparison to Simple or Antithetic Variates sampling.

Development and laboratory evaluation of a compact swirling aerosol sampler (SAS) for collection of atmospheric bioaerosols

Inhalation of infectious bioaerosols has been linked to a variety of respiratory diseases. However, efficient sampling techniques to allow high temporal resolution sampling are limited to collect and study bioaerosols in the various occupational and ambient micro–environmental atmospheres. This study introduces a medium flow swirling bioaerosol sampler (SAS) approach that collects atmospheric bioaerosols at the flow rate of 167 Lpm (10 cubic meter per hour). The collection of bioaerosols is achieved through a combination of impaction and cyclonic centrifugal motion. Aerosol deposition efficiency tests were performed with monodispersive polystyrene latex (PSL) particles ranging from 0.1 to 10 μm. Results have shown that the sampler has cut–off size of 0.7 μm and 1.5 μm, with and without the assistance of added water vapor, respectively. The bioaerosol collection and viability tests were performed with comparison to the commercially–available BioSampler, and the results show that the collection efficiency of the SAS is 97% at the designed flow rate, while the higher flow of the new system yields more than 13 times of the collection rate compared to the BioSampler. The high collection efficiency and observed viability preservation of the SAS make it an attractive alternative for high time resolution bioaerosol sampling for atmospheric, occupational and indoor air quality monitoring.

Likelihood Based Inference and Prediction in Spatio-Temporal Panel Count Models for Urban Crimes

We develop a panel data count model combined with a latent Gaussian spatio-temporal heterogenous state process to analyze monthly severe crimes at the census tract level in Pittsburgh, Pennsylvania. Our data set combines Uniform Crime Reporting data with socio-economic data from the 2000 census. The likelihood of the model is accurately estimated by adapting recently developed efficient importance sampling techniques applicable to high-dimensional spatial models with sparse precision matrices. Our estimation results confirm socio-economic explanations for crime and, foremost, the broken-windows hypothesis, whereby less severe crimes in a region is a leading indicator for severe crimes. In addition to ML parameter estimates, we compute several other statistics of interest for law enforcement such as elasticities (idiosyncratic, total, short-term as well as long-term) of severe crimes w.r.t. less severe crimes, one-month-ahead out-of-sample forecasts, predictive cumulative distribution functions and validation test statistics based on these cdf's.

On the Pareto optimality of variance reduction simulation techniques in structural reliability

Estimation of low failure probabilities in high dimensional structural reliability problems generally involves a trade-off between computational effort and accuracy of the estimate, whether efficient sampling techniques have been employed or not. While a substantial effort continues to be made by the community to develop and benchmark new and efficient sampling schemes, the limits of performance of a given algorithm, e.g., what is the best attainable accuracy of the method for a fixed computational effort and if that is good enough, have not received comparable attention. However, such insights could prove valuable in making the right choice in solving a computationally demanding reliability problem. In a multi-objective stochastic optimization formulation, these questions yield the so-called Pareto front or the set of non-dominated solutions: solutions that cannot be further improved without worsening at least one objective. Posteriori user defined preferences can then be applied to rank members of the Pareto set and obtain the best strategy. We take up two classes of variance-reducing algorithms – importance sampling (IS) and subset simulations (SS) – and apply them to a range of benchmarked reliability problems of various size and complexity to bring out the issue of optimality and trade-off between accuracy and effort. The design variables are variously of categorical, discrete as well as continuous types and the stochastic multi-objective optimization without recourse is solved using Genetic Algorithms. In each case, we ascertain the best possible accuracies that a given method can achieve and identify the corresponding design variables. We find that the proposal pdf does have an effect on the efficiency of SS, the FORM design point is not always the best sampling location in IS and setting the sensitivity parameter associated with Adaptive Importance Sampling at 0.5 does not guarantee optimal performance. In addition to this the benefits of using SS for high dimensional problems are reinforced. We also show that the Pareto fronts corresponding to different methods can intersect indicating that more is not always better and different solution techniques for the same problem may be required in different computational regimes.

Recursive Bayesian filtering in circular state spaces

For recursive circular filtering based on circular statistics, we introduce a general framework for estimation of a circular state based on different circular distributions, specifically the wrapped normal distribution and the von Mises distribution. We propose an estimation method for circular systems with nonlinear system and measurement functions. This is achieved by relying on efficient deterministic sampling techniques. Furthermore, we show how the calculations can be simplified in a variety of important special cases, such as systems with additive noise as well as identity system or measurement functions. We introduce several novel key components, particularly a distribution-free prediction algorithm, a new and superior formula for the multiplication of wrapped normal densities, and the ability to deal with non-additive system noise. All proposed methods are thoroughly evaluated and compared to several state-of-the-art solutions.

Image statistics underlying natural texture selectivity of neurons in macaque V4

Our daily visual experiences are inevitably linked to recognizing the rich variety of textures. However, how the brain encodes and differentiates a plethora of natural textures remains poorly understood. Here, we show that many neurons in macaque V4 selectively encode sparse combinations of higher-order image statistics to represent natural textures. We systematically explored neural selectivity in a high-dimensional texture space by combining texture synthesis and efficient-sampling techniques. This yielded parameterized models for individual texture-selective neurons. The models provided parsimonious but powerful predictors for each neuron's preferred textures using a sparse combination of image statistics. As a whole population, the neuronal tuning was distributed in a way suitable for categorizing textures and quantitatively predicts human ability to discriminate textures. Together, we suggest that the collective representation of visual image statistics in V4 plays a key role in organizing the natural texture perception.

A triple-threshold leverage stochastic volatility model

Abstract In this paper we introduce a triple-threshold leverage stochastic volatility (TTLSV) model for financial return time series. The main feature of the model is to allow asymmetries in the leverage effect as well as mean and volatility. In the model the asymmetric effect is modeled by a threshold nonlinear structure that the two regimes are determined by the sign of the past returns. The model parameters are estimated using maximum likelihood (ML) method based on the efficient importance sampling (EIS) technique. Monte Carlo simulations are presented to examine the accuracy and finite sample properties of the proposed methodology. The EIS-based ML (EIS-ML) method shows good performance according to the Monte Carlo results. The proposed model and methodology are applied to two stock market indices for China. Strong evidence of the mean and volatility asymmetries is detected in Chinese stock market. Moreover, asymmetries in the volatility persistence and leverage effect are also discovered. The log-likelihood and Akaike information criterion (AIC) suggest evidence in favor of the proposed model. In addition, model diagnostics suggest that the proposed model performs relatively well in capturing the key features of the data. Finally, we compare models in a Value at Risk (VaR) study. The results show that the proposed model can yield more accurate VaR estimates than the alternatives.

Molecular simulation of water and hydration effects in different environments: challenges and developments for DFTB based models.

We discuss the description of water and hydration effects that employs an approximate density functional theory, DFTB3, in either a full QM or QM/MM framework. The goal is to explore, with the current formulation of DFTB3, the performance of this method for treating water in different chemical environments, the magnitude and nature of changes required to improve its performance, and factors that dictate its applicability to reactions in the condensed phase in a QM/MM framework. A relatively minor change (on the scale of kBT) in the O–H repulsive potential is observed to substantially improve the structural properties of bulk water under ambient conditions; modest improvements are also seen in dynamic properties of bulk water. This simple change also improves the description of protonated water clusters, a solvated proton, and to a more limited degree, a solvated hydroxide. By comparing results from DFTB3 models that differ in the description of water, we confirm that proton transfer energetics are adequately described by the standard DFTB3/3OB model for meaningful mechanistic analyses. For QM/MM applications, a robust parametrization of QM-MM interactions requires an explicit consideration of condensed phase properties, for which an efficient sampling technique was developed recently and is reviewed here. The discussions help make clear the value and limitations of DFTB3 based simulations, as well as the developments needed to further improve the accuracy and transferability of the methodology.

INDICADORES DE DIVERSIDAD, ESTRUCTURA Y RIQUEZA PARA LA CONSERVACIÓN DE LA BIODIVERSIDAD VEGETAL EN LOS PAISAJES RURALES

<p>The lack of efficient sampling techniques for the study of linear vegetation remnants in the Mexican tropics anthropised areas was the main premise for this research, in order to determine the diversity, richness and vegetation structure of rural landscapes according to its conservation value through efficient and representative linear sampling techniques. The biological characterization of plants was performed in an area of 4,096 m<sup>2</sup>, in eight components of the landscape, a total of 2260 ha. The vegetation consists of 144 species, 127 genera and 51 families. These were distributed at herbs 71.26%, shrubs 22.80%, trees 3.49%, and vines 2.45%. High values of diversity and richness were found in hedgerows, stream shore, shore road, scrub and citrus growing. It is suggested to give priority to the conservation of stream bank that protects species such as <em>Cedrela</em> <em>odorata</em> and <em>Sapium macrocarpum</em>. The efficiency and representativeness of the sample was 95% through rarefaction given by the best fit of Chao 1 and ACE, highly significant linear correlation between true diversity and richness of species, and the high value of spare.</p>

Maximum likelihood estimation of partially observed diffusion models

This paper develops a maximum likelihood (ML) method to estimate partially observed diffusion models based on data sampled at discrete times. The method combines two techniques recently proposed in the literature in two separate steps. In the first step, the closed form approach of Aït-Sahalia (2008) is used to obtain a highly accurate approximation to the joint transition probability density of the latent and the observed states. In the second step, the efficient importance sampling technique of Richard and Zhang (2007) is used to integrate out the latent states, thereby yielding the likelihood function. Using both simulated and real data, we show that the proposed ML method works better than alternative methods. The new method does not require the underlying diffusion to have an affine structure and does not involve infill simulations. Therefore, the method has a wide range of applicability and its computational cost is moderate.

An efficient optimum Latin hypercube sampling technique based on sequencing optimisation using simulated annealing

This paper proposes a new optimal Latin hypercube sampling method (OLHS) for design of a computer experiment. The new method is based on solving sequencing and continuous optimisation using simulated annealing. There are two sets of design variables used in the optimisation process: sequencing and real number variables. The special mutation operator is developed to deal with such design variables. The performance of the proposed numerical strategy is tested and compared with three established OLHS methods, namely genetic algorithm (GA), enhanced stochastic evolutionary algorithm (ESEA) and successive local enumeration (SLE). Based on 30 test problems with various design dimensions and numbers of sampling points, the proposed method gives the best results. The method can generate an optimum set of sampling points within reasonable computing time; therefore, it can be considered as a powerful tool for design of computer experiments.

A sampling algebra for aggregate estimation

As of 2005, sampling has been incorporated in all major database systems. While efficient sampling techniques are realizable, determining the accuracy of an estimate obtained from the sample is still an unresolved problem. In this paper, we present a theoretical framework that allows an elegant treatment of the problem. We base our work on generalized uniform sampling (GUS), a class of sampling methods that subsumes a wide variety of sampling techniques. We introduce a key notion of equivalence that allows GUS sampling operators to commute with selection and join, and derivation of confidence intervals. We illustrate the theory through extensive examples and give indications on how to use it to provide meaningful estimates in database systems.

Correcting the nondetection bias of angle count sampling

The well-known angle count sampling (ACS) has proved to be an efficient sampling technique and has been applied in forest inventories for many decades. However, ACS assumes total visibility of objects; any violation of this assumption leads to a nondetection bias. We present a novel approach, in which the theory of distance sampling is adapted to traditional ACS to correct for the nondetection bias. Two new estimators were developed based on expanding design-based inclusion probabilities by model-based estimates of the detection probabilities. The new estimators were evaluated in a simulation study as well as in a real forest inventory. It is shown that the nondetection bias of the traditional estimator is up to −52.5%, whereas the new estimators are approximately unbiased.

Preparation of novel alumina nanowire solid-phase microextraction fiber coating for ultra-selective determination of volatile esters and alcohols from complicated food samples

A novel alumina nanowire (ANW) solid-phase microextraction (SPME) fiber coating was prepared by a simple and rapid anodization-chemical etching method for ultra-selective determination of volatile esters and alcohols from complicated food samples. Preparation conditions for ANW SPME fiber coating including corrosion solution concentration and corrosion time were optimized in detail for better surface morphology and higher surface area based on scanning electron microscope (SEM). Under the optimum conditions, homogeneous alumina nanowire structure of ANW SPME fiber coating was achieved with the average thickness of 20μm around. Compared with most of commercial SPME fiber coatings, ANW SPME fiber coatings achieved the higher extraction capacity and special selectivity for volatile esters and alcohols. Finally, an efficient gas sampling technique based on ANW SPME fiber coating as the core was established and successfully applied for the ultra-selective determination of trace volatile esters and alcohols from complicated banana and fermented glutinous rice samples coupled with gas chromatography/mass spectrometry (GC/MS) detection. It was interesting that 25 esters and 2 alcohols among 30 banana volatile organic compounds (VOCs) identified and 4 esters and 7 alcohols among 13 identified VOCs of fermented glutinous rice were selectively sampled by ANW SPME fiber coatings. Furthermore, new analytical methods for the determination of some typical volatile esters and alcohols from banana and fermented glutinous rice samples at specific storage or brewing phases were developed and validated. Good recoveries for banana and fermented glutinous rice samples were achieved in range of 108–115% with relative standard deviations (RSDs) of 2.6–6.7% and 80.0–91.8% with RSDs of 0.3–1.3% (n=3), respectively. This work proposed a novel and efficient gas sampling technique of ANW SPME which was quite suitable for ultra-selectively sampling trace volatile esters and alcohols from complicated food samples.