Search

Markov chain Monte Carlo (MCMC) is an essential set of tools for estimating features of probability distributions commonly encountered in modern applications. For MCMC simulation to produce reliable outcomes, it needs to generate observations representative of the target distribution, and it must be long enough so that the errors of Monte Carlo estimates are small. We review methods for assessing the reliability of the simulation effort, with an emphasis on those most useful in practically relevant settings. Both strengths and weaknesses of these methods are discussed. The methods are illustrated in several examples and in a detailed case study. Expected final online publication date for the Annual Review of Statistics, Volume 9 is March 2022. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.

Numerous studies have shown that near repeat victimization of burglaries can account for a substantial minority of burglaries in urban settings. Using a method based on the distribution of potential targets to determine the size of spatial bandwidths, the presence of burglary near repeats in rural areas was examined and compared to the level of near repeats in urban areas. A significant burglary near repeat pattern was observed in rural areas, but was restricted to the spatial and temporal bands that were closest to and most recently after a previous burglary. The proportion of all burglaries that were near repeats in this nearest spatial and temporal bands was greater than that observed in urban areas. The findings lead to extending how the boost account and offender foraging principles may apply in rural settings, and the identification of crime prevention opportunities that counter near repeats in rural areas.

We consider the problem of sampling from a target distribution, which is not necessarily log-concave, in the context of empirical risk minimization and stochastic optimization as presented in [M. R...

In this note we construct an algorithm generating any discrete distribution with an arbitrary coin (and, as a result, with arbitrary initial distribution). The coin need not be fair and the target distribution can be supported on a countable set.

The experience of modern armed conflicts and exercises shows the need for a detailed assessment of enemy targets, especially during the planning and conduct of raids. The article considers approaches to assessing the importance of enemy objects, which are based on the calculation of combat potentials and determining the coefficients of operational and tactical importance of enemy objects. The authors identify one of the main problems - the lack of a common understanding of the order of determining the importance (priority) of enemy objects, especially when planning raids. The purpose of this article is to consider the application of the method of hierarchy analysis to determine the coefficients of operational and tactical importance of enemy objects during the planning and conduct of raids. The authors disclose the general provisions and algorithm of decision-making using this method. To better understand the procedure for applying the method of analysis of hierarchies, an example of forming a hierarchical image of the problem, the procedure for forming matrices of pairwise comparisons depending on the level of hierarchy and their completion, mathematical processing of results, the procedure for checking expert judgments and the whole hierarchy. It is noted that this technique is part of a mathematical model of target distribution of various types of fire on enemy targets. The application of the technique will allow to determine the possibilities of the raid detachment to destroy (disable) enemy objects.

Quantile matching is a strictly monotone transformation that sends the observed response values $\{y_1, . . . , y_n\}$ to the quantiles of a given target distribution. A likelihood based criterion is developed for comparing one target distribution with another in a linear-model setting., 8 pages, 2 figures

MCMC algorithm is well-known for having difficulty exploring distant modes when the target distribution is multi-modal. The reason is that a proposal state is likely to be rejected when traversing ...

The Monte Carlo within Metropolis (MCwM) algorithm, interpreted as a perturbed Metropolis-Hastings (MH) algorithm, provides an approach for approximate sampling when the target distribution is intractable. Assuming the unperturbed Markov chain is geometrically ergodic, we show explicit estimates of the\ud difference between the n-th step distributions of the perturbed MCwM and the\ud unperturbed MH chains. These bounds are based on novel perturbation results\ud for Markov chains which are of interest beyond the MCwM setting. To apply the\ud bounds, we need to control the difference between the transition probabilities\ud of the two chains and to verify stability of the perturbed chain.\ud Keywords: Markov chain Monte Carlo, restricted approximation, Monte\ud Carlo within Metropolis, intractable likelihood.

Many proteins that undergo sequential enzymatic modification in the Golgi cisternae are displayed at the plasma membrane as cell identity markers. The modified proteins, called glycans, represent a molecular code. The fidelity of this glycan code is measured by how accurately the glycan synthesis machinery realises the desired target glycan distribution for a particular cell type and niche. In this paper, we quantitatively analyse the tradeoffs between the number of cisternae and the number and specificity of enzymes, in order to synthesize a prescribed target glycan distribution of a certain complexity. We find that to synthesize complex distributions, such as those observed in real cells, one needs to have multiple cisternae and precise enzyme partitioning in the Golgi. Additionally, for fixed number of enzymes and cisternae, there is an optimal level of specificity of enzymes that achieves the target distribution with high fidelity. Our results show how the complexity of the target glycan distribution places functional constraints on the Golgi cisternal number and enzyme specificity., Comment: 30 pages

We introduce a generalized version of the Metropolis-adjusted Langevin algorithm (MALA). The informed proposal distribution of this new sampler features two tuning parameters: the usual step size parameter $$\sigma ^2$$ and an interpolation parameter $$\gamma $$ that may be adjusted to accommodate the dimension of the target distribution. We theoretically study the efficiency of the sampler by making use of the local- and global-balance concepts introduced in Zanella (JASA 115:852–865, 2020) and provide efficient tuning guidelines that work well with a variety of target distributions. Although the usual MALA ( $$\gamma =1$$ ) is shown to be optimal for infinite-dimensional targets, in practice, the generalized MALA ( $$1

Background: Little is known about the differences among ventricular arrhythmias (VAs) ablated in different subregions of the aortic sinuses of Valsalva (ASVs). We aim to investigate the distribution, precordial electrocardiographic patterns, and bipolar electrogram characteristics of VAs ablated in different subregions of the ASVs.Methods: We divided the right ASV and the left ASV into a total of 6 subregions and studied 51 idiopathic VAs ablated first time successfully in the ASVs.Results: These 51 VAs were inhomogeneously distributed among the 6 subregions, which comprised the right-lateral ASV (1/51), the right-anterior ASV (11/51), the regions along the right (13/51) and left (9/51) sides of the ASV junction, the left-anterior ASV (5/51), and the left-lateral ASV (12/51). Fractionated potentials were dominant (39/51, 76%) among the 3 types of target electrograms. From the right-lateral ASV to the left-lateral ASV, the percentage of fractionated potentials gradually decreased from 100 to 59%. A precordial rebound notch in V3-V4 or V4-V5 had sensitivity of 90.9%, specificity of 85.0%, and negative predictive value (NPV) of 97.1% to predict VAs ablated in the right-anterior ASV. A precordial rebound notch in V2-V3 had sensitivity of 50.0%, specificity of 94.9%, and NPV of 86.0% to predict VAs ablated in the left-lateral ASV.Conclusion: VA targets were mainly distributed in the anterior and the left-lateral ASVs. Fractionated potentials were common among target electrograms, especially in theright-anterolateral ASV. Precordial electrocardiographic rebound notch has high predictive accuracy in identifying different subregions of the ASVs as target ablation sites.

Planar array electrical capacitance tomography (ECT) has an attractive prospect for underground detection of non-metallic materials, e.g., landmine and subsurface cavity detection. Accurate 3D image reconstruction is essential to these applications. This paper presents a 3D reconstruction method based on depth estimation and sparse representation. The approximate depth of object is attained by postprocessing an image by Tikhonov method and regarded as the prior knowledge for the sparse representation of target distribution. The 3D reconstruction is achieved by solving a dedicated optimization problem using the Gauss-Newton method. To simulate the application of subsurface cavity detection, a planar array ECT model has been established and solved numerically. With five targets of various shapes and depths, the estimated depths have been calculated and compared. Those five targets various shapes are then reconstructed using the proposed method and conventional methods, with their performances evaluated and compared. The preliminary results validated the superior performance of the proposed method.

We introduce a new variational inference (VI) framework, called energetic variational inference (EVI). It minimizes the VI objective function based on a prescribed energy-dissipation law. Using the EVI framework, we can derive many existing Particle-based Variational Inference (ParVI) methods, including the popular Stein Variational Gradient Descent (SVGD) approach. More importantly, many new ParVI schemes can be created under this framework. For illustration, we propose a new particle-based EVI scheme, which performs the particle-based approximation of the density first and then uses the approximated density in the variational procedure, or "Approximation-then-Variation" for short. Thanks to this order of approximation and variation, the new scheme can maintain the variational structure at the particle level, and can significantly decrease the KL-divergence in each iteration. Numerical experiments show the proposed method outperforms some existing ParVI methods in terms of fidelity to the target distribution., Comment: 17 pages, 7 figures

We correct an error in Theorem 2.1 in Bladt and Sørensen (Bernoulli 20 (2014) 645–675), where the initial distribution of an auxiliary diffusion process that is used to describe the distribution of the proposed approximate diffusion bridge is wrong. As a consequence, we also correct the pseudo marginal Metropolis-Hastings algorithm that has an exact diffusion bridge as its target distribution. The same auxiliary diffusion plays a central role in the algorithm.

Short text clustering is challenging in the field of Natural Language Processing (NLP) since it is hard to learn the discriminative representations with limited information. In this paper, fused multi-embedded features are employed to enhance the representations of short texts. Then, a denoising autoencoder with an attention layer is adopted to extract low-dimensional features from the multi-embeddings against the disturbance of noisy texts. Furthermore, we propose a novel distribution estimation with jointly utilizing soft cluster assignment and the prior target distribution transition to better fine-tune the encoder. Combining the above work, we propose a deep multi-embedded self-supervised model(DMESSM) for short text clustering. We compare our DMESSM with the state-of-the-art methods in head-to-head comparisons on benchmark datasets, which indicates that our method outperforms them.

The experimental design for a generalized linear model (GLM) is important but challenging since the design criterion often depends on model specification including the link function, the linear predictor, and the unknown regression coefficients. Prior to constructing locally or globally optimal designs, a pilot experiment is usually conducted to provide some insights on the model specifications. In pilot experiments, little information on the model specification of GLM is available. Surprisingly, there is very limited research on the design of pilot experiments for GLMs. In this work, we obtain some theoretical understanding of the design efficiency in pilot experiments for GLMs. Guided by the theory, we propose to adopt a low-discrepancy design with respect to some target distribution for pilot experiments. The performance of the proposed design is assessed through several numerical examples.

Common domain shift problem formulations consider the integration of multiple source domains, or the target domain during training. Regarding the generalization of machine learning models between different car interiors, we formulate the criterion of training in a single vehicle: without access to the target distribution of the vehicle the model would be deployed to, neither with access to multiple vehicles during training. We performed an investigation on the SVIRO dataset for occupant classification on the rear bench and propose an autoencoder based approach to improve the transferability. The autoencoder is on par with commonly used classification models when trained from scratch and sometimes out-performs models pre-trained on a large amount of data. Moreover, the autoencoder can transform images from unknown vehicles into the vehicle it was trained on. These results are corroborated by an evaluation on real infrared images from two vehicle interiors., Comment: This paper has been accepted at IEEE Intelligent Vehicles Symposium (IV), 2021

We consider the problem of search track placement over a rectangular region that considers a priori target information to maximize coverage. Our approach allocates uniformly spaced parallel tracks over sections of homogeneous sensor performance. We solve the search planning problem as a MILP optimization problem that is subject to a priori target distribution constraints, as well as minimum sectional and total effective coverage constraints. Additionally, we provide a derivation of a general method to evaluate the effective coverage that remains agnostic to the assumption of dependence or independence between tracks, as well as account for the inevitable overlap of parallel tracks. We include simulation results that demonstrate the efficacy of our approach.

We propose an efficient algorithm to learn the transition probabilities of a Markov chain in a way that its weighted PageRank scores meet some predefined target values. Our algorithm does not require any additional information about the nodes and the edges in the form of features, i.e., it solely considers the network topology for calibrating the transition probabilities of the Markov chain for obtaining the desired PageRank scores. Our experiments reveal that we can reliably and efficiently approximate the probabilities of the transition matrix, resulting in the weighted PageRank scores of the nodes to closely match some target distribution. We demonstrate our findings on both quantitative and qualitative evaluations by reporting experimental results on web traffic (the English Wikipedia and a Hungarian news portal) and the bicycle sharing network of New York City.

Decades of research have provided evidence that object representations contribute to attentional selection. However, most evidence for object-based attentional allocation is drawn from studies employing the two-rectangle paradigm where the target distribution is biased towards the cued object. It is thus unclear whether object-based attentional selection is from object representations or a consequence of spatial attention based on statistical imbalances. Here, we investigate the extent to which target frequency modulates object-based attention by systematically manipulating the frequency of target appearance in a particular spatial location within objects to equate spatial allocation, bias specific spatial locations, or bias objects. In four experiments, participants were presented with a variant of the two-rectangle paradigm in which one end of a rectangle was cued and performed a target discrimination task. Critically, the target location probabilities were parametrically manipulated. The target could appear equally in all ends within the objects (valid, invalid within-object, invalid between-object, diagonal) (Experiment 1) or with overall equality between objects but biased towards the invalid locations (Experiment 2). The target could also appear in three locations (valid, invalid within-object, invalid between-object) distributed equally between objects but biased towards the invalid between-object location (Experiment 3) or with an overall bias towards the invalid between-object location (Experiment 4). We observed that while objects bias attention, spatial biases are prioritized over object representations. Combined results suggest that object-based contribution to attentional guidance is the result of both spatial probabilities and object representations.

To achieve handy and reliable quantitative measurements of surface-enhanced Raman spectroscopy (SERS) for universal analysis, this work designs a self-calibrated 2D nanoarchitecture as a flexible sensor. This sensor consists of a self-assembled gold nanostars (AuNSs) monolayer sandwiched between two pieces of graphene, and can be easily attached on the surface of arbitrary samples. The inner AuNSs can offer a huge number of hot spots for strong Raman enhancement, and the existence of graphene further improves the stability and reproducibility of SERS signals for precise detection. By implanting Raman reporters as internal standard (IS) onto one side of the device and attaching the device with opposite side on sample surface, the distinct SERS signals from both IS and target can be quickly collected, which offers reliable quantification information of the target by IS calibration and has been demonstrated at the theoretical and experimental levels. Using carcinogenic dyestuff and agricultural fungicide as target models, this flexible nanoarchitecture realizes fast quantitative response to target in sample solution, and rapid in situ imaging of target distribution on solid surface. This sensor shows excellent reusability, good durability, and diversified structure collocation along with the simple operability, indicating great promise for fast on-site SERS detection in quantification and imaging applications.

We propose asset and liability management (ALM) models in which the risk of underfunding is modelled based on the concept of stochastic dominance. Investment decisions are taken such that the distribution of the funding ratio, that is, the ratio of asset to liabilities, is non-dominated with respect to second order stochastic dominance (SSD). In addition, the funding ratio distribution is close in an optimal sense to a user-specified target distribution. Interesting results are obtained when the target distribution is degenerate; in this case, we can obtain equivalent risk minimisation models, with risk defined as expected shortfall or as worst case loss. As an application, we consider the financial planning problem of a defined benefit pension fund in Saudi Arabia.

This study provides a general framework for the creation of three-dimensional microstructures for anisotropic two-phase materials with complex morphologies. The ability to create synthetic microstructures that are statistically representative of real materials is dependent on the ability of the generation algorithm to meet the assigned statistics. Hence, the fidelity of the synthetic microstructure depends on the efficiency of the generation process. The generated microstructures were statistically evaluated with respect to the target distribution for grain size, shape, orientation and phase volume fraction. This serves a dual purpose of validation of the packing algorithm and quantification of the 3D microstructures.

Ethnopharmacological relevance Precise target distribution is a key issue for further appropriate applications of fangjis (formulas) with similar efficacy and herbal constituents to maximize efficacy and minimize toxicity. Aim of the study To develop an algorithm for investigating the differential target distributions and characteristic mechanisms of fangjis. Main methods In this study, we proposed a Modulome-Fangjiome Association Study (MoFAS) for comparing fangjis from qi-invigorating and xue-nourishing fangjiome (represented by four fangjis: FEJ, SDT, LYG and QOL). Firstly, the database-driven target network of these 4 fangjis was constructed as qi-xue network and decomposed into modules. Then, the modular map with functional landscape were constructed based on consistency score and enrichment analysis. Finally, we employed a targeting rate (TR) matrix to assess the contribution of this fangjiome to modulome (a set of modules) and compared characteristic effect of fangjis by principal component analysis (PCA). Results A qi-xue network constituted by 579 proteins and 23 modules were constructed. In the functional landscape, 3 primary modules were mainly involved in the endocrine system and environmental adaptation. For the target distribution, SDT and QOL were more similar; the FEJ and LYG were located distant from other fangjis according to PCA. The common effects of FEJ, SDT, and QOL focused on stress response and organism development in environmental perturbation, but the FEJ was superior in regulating critical targets, primarily focusing on hormone and neurotransmitter processes. SDT and QOL were concentrated on the majority scale of the qi-xue network, especially for the mitotic cell cycle and development. LYG only targeted lymphocyte costimulation and icosanoid biosynthetic processes. Conclusion In this study, for the first time, we investigated the difference in the target distribution of qi-invigorating and xue-nourishing fangjiome and provided direct evidence for the characteristic therapeutic effect of these fangjis, which may promote the precise application of fangjis and support the identification of appropriate populations.

When making risky spatial decisions, humans incorporate estimates of sensorimotor variability and costs on outcomes to bias their spatial selections away from regions that incur feedback penalties. Since selection variability depends on the reliability of sensory signals, increasing the spatial variance of targets during visually guided actions should increase the degree of this avoidance. Healthy adult participants (N = 20) used a computer mouse to indicate their selection of the mean of a target, represented as a 2D Gaussian distribution of dots presented on a computer display. Reward feedback on each trial corresponded to the estimation error of the selection. Either increasing or decreasing the spatial variance of the dots modulated the spatial uncertainty of the target. A non-target distractor cue was presented as an adjacent distribution of dots. On a subset of trials, feedback scores were penalized with increased proximity to the distractor mean. As expected, increasing the spatial variance of the target distribution increased selection variability. More importantly, on trials where proximity to the distractor cue incurred a penalty, increasing variance of the target increased selection bias away from the distractor cue and prolonged reaction times. These results confirm predictions that increased sensory uncertainty increases avoidance during risky spatial decisions.

The authors study an optimal searcher path problem of the following type. A team of searchers tries to find an immobile target on a unit network (in a unit network, all edges are of unit length). The target is located at some point on this network, and its location probability distribution is known to the searcher team. Searchers move along the edges on the network at a constant speed. When one searcher passes the target location, detection happens and the search ends. The objective is to find paths for this team of searchers to minimize the expected search time. A linear binary programming model is built to find the optimal search strategy. Computational experiments are conducted to test the capability of this model, and to analyze how searcher team size and target distribution might affect the search time. Analytical results are given for a special problem scenario: one searcher searching for a uniformly distributed target on a tree. We prove that, any depth-first search path is optimal for a given searcher start point, and the best start point is one of the leaf nodes.

The multiset sampler has been shown to be an effective algorithm to sample from complex multimodal distributions, but the multiset sampler requires that the parameters in the target distribution can be divided into two parts: the parameters of interest and the nuisance parameters. We propose a new self-multiset sampler (SMSS), which extends the multiset sampler to distributions without nuisance parameters. We also generalize our method to distributions with unbounded or infinite support. Numerical results show that the SMSS and its generalization have a substantial advantage in sampling multimodal distributions compared to the ordinary Markov chain Monte Carlo algorithm and some popular variants. Supplemental materials for the article are available online.

We present a diagnostic for monitoring convergence of a Markov chain Monte Carlo (MCMC) sampler to its target distribution. In contrast to popular existing methods, we monitor convergence to the joint target distribution directly rather than a select scalar projection. The method uses a simple nonparametric posterior approximation based on a state-space partition obtained by clustering the pooled draws from multiple chains, and convergence is determined when the estimated posterior probabilities of partition elements under each chain are sufficiently similar. This framework applies to a wide variety of problems, and generalizes directly to non-Euclidean state spaces. Our method also provides approximate high-posterior-density regions, and a characterization of differences between nonconverged chains, all with little additional computational burden. We demonstrate this approach on applications to sampling posterior distributions over Rp, graphs, and partitions. Supplementary materials for this arti...

This paper considers two different control problems for deterministic systems with stochastic initial conditions where, in addition to the usual asymptotic behavior requirement, we are interested in the transient behavior of the state distribution evolution. For the first one, we study control design such that the state trajectories enter target set at a given transient time with a prescribed minimum cumulative distribution. For the second one, we propose control design where the distribution of state variable at transient time is close to the target distribution. We illustrate the efficacy of the proposed solutions to the aforementioned control problems through numerical examples.

The understanding of metabolic disease and diabetes on a molecular level has increased significantly due to the recent advances in molecular biology and biotechnology. However, in vitro studies and animal models do not always translate to the human disease, perhaps illustrated by the failure of many drug candidates in the clinical phase. Non-invasive biomedical imaging techniques such as Positron Emission Tomography (PET) offer tools for direct visualization and quantification of molecular processes in humans. Developments in this area potentially enable longitudinal in vivo studies of receptors and processes involved in diabetes guiding drug development and diagnosis in the near future. This mini-review focuses on describing the overall perspective of how PET can be used to increase our understanding and improve treatment of diabetes. The methodological aspects and future developments and challenges are highlighted.

A complementary to SMC framework for effective sampling in high-dimensions is that of Markov chain Monte Carlo (MCMC). In this framework there is a single target distribution but samples are drawn sequentially as a Markov chain that is invariant with respect to the target. For the purposes of this book we are primarily interested in the MCMC kernels that are used to generate the successive samples because we can use them as transition kernels in Feynman-Kac models. The combination of SMC with MCMC kernels, which we develop in subsequent chapters, creates opportunities for optimising parameters in the MCMC kernels using information about the target extracted from the particle population. This chapter lays down the fundamental notions, the most important of which for the purposes of this book being the notion of invariance, and a few typical MCMC kernels.

A deep generative model such as a GAN learns to model a rich set of semantic and physical rules about the target distribution, but up to now, it has been obscure how such rules are encoded in the network, or how a rule could be changed. In this paper, we introduce a new problem setting: manipulation of specific rules encoded by a deep generative model. To address the problem, we propose a formulation in which the desired rule is changed by manipulating a layer of a deep network as a linear associative memory. We derive an algorithm for modifying one entry of the associative memory, and we demonstrate that several interesting structural rules can be located and modified within the layers of state-of-the-art generative models. We present a user interface to enable users to interactively change the rules of a generative model to achieve desired effects, and we show several proof-of-concept applications. Finally, results on multiple datasets demonstrate the advantage of our method against standard fine-tuning methods and edit transfer algorithms.

We investigate and characterize the inherent resilience of conditional Generative Adversarial Networks (cGANs) against noise in their conditioning labels, and exploit this fact in the context of Unsupervised Domain Adaptation (UDA). In UDA, a classifier trained on the labelled source set can be used to infer pseudo-labels on the unlabelled target set. However, this will result in a significant amount of misclassified examples (due to the well-known domain shift issue), which can be interpreted as noise injection in the ground-truth labels for the target set. We show that cGANs are, to some extent, robust against such "shift noise". Indeed, cGANs trained with noisy pseudo-labels, are able to filter such noise and generate cleaner target samples. We exploit this finding in an iterative procedure where a generative model and a classifier are jointly trained: in turn, the generator allows to sample cleaner data from the target distribution, and the classifier allows to associate better labels to target samples, progressively refining target pseudo-labels. Results on common benchmarks show that our method performs better or comparably with the unsupervised domain adaptation state of the art.

Obtaining good samples from high dimensional target distributions by adaptive importance sampling (AIS) methods to approximate target distributions is usually very hard, especially when not much is known about the target distribution. The main difficulty is in adapting the proposal distribution to become closer to the target distribution. The Variational Auto-Encoder (VAE) is a probabilistic deep computational network that can capture lower dimensional coding of high dimensional data. This property gives it a potential to capture the structure of high dimensional target distributions. In this paper, we propose an AIS method that exploits variational autoencoding to construct better proposal distributions. We demonstrate that the proposed method has better performance than the standard AIS in terms of convergence speed and goodness-of-fit when sampling from high dimensional and highly structured target distributions.

BACKGROUND Naloxone is an opioid antagonist that can reverse an opioid overdose. Increased opioid-related mortality rates led to greater distribution of naloxone without a prescription and administration of naloxone by laypersons. This study fills a gap in knowledge of naloxone experiences among active users of opioids living in suburban communities. PURPOSE The purpose of this article is to provide nurse practitioners with an in-depth understanding of current naloxone use practices among people who experience overdose events. The specific aims are to compare access to naloxone in diverse suburban towns, to examine administration differences across settings, and to understand perspectives on naloxone experiences from people who are actively using opioids. METHODOLOGICAL ORIENTATION The data for this analysis were drawn from an ethnographic study in the suburban towns around Atlanta, Georgia; Boston, Massachusetts; and New Haven, Connecticut. Short surveys and in-depth interviews were collected. Inductive methods were used to compare data across settings. SAMPLE The sample of 106 included 48% female, 62% White, 24% African American/Black, 13% more than one race, and 21% Hispanic/Latinx. The mean age was 41.35 years. CONCLUSIONS Differences between study settings in access to naloxone, administration frequency, and delivery systems were found. Findings suggest more education and training is needed in overdose prevention and harm reduction intervention. Studies on delivery systems need to address the increase in fentanyl-related overdoses. IMPLICATIONS FOR PRACTICE Nurse practitioners can help to target distribution of naloxone in local communities, facilitate collaboration with harm reduction services, and provide evidence-based education and training to laypersons.

Shaping field intensity in target areas is a challenging problem in electromagnetics. Most of the existing methods can realize a certain type of field like plane wave field and focusing field. In this paper, we present an array design and control approach for shaping the field intensity to different patterns without dependence on precise control of phase. At the same time, the excitation parameters of array can be theoretical calculated according to target distribution of field intensity. We have studied the tricorn distributions in a 1-D target area. The results of theoretical calculation and full-wave simulation demonstrate the effectiveness of the proposed method.

Mobile crowd sensing (MCS) collects city-scale sensing data with low cost and high efficiency. One important goal of MCS is to ensure high quality of sensing coverage to provide sufficient information to data analysis end. However, the goal of the MCS may be inconsistent with the goal of vehicles. This inconsistency between goals results in a bad sensing coverage and decreases the quality of the collected information. Key challenges to resolve this inconsistency include the heterogeneous target desired spatio-temporal distributions, limited budget constraining the ability to incentivize more taxis, and high computational complexity. This work shows a vehicle dispatching system to optimize the sensing coverage of the sampled data with a limited budget and taxis. Our system models the objective function using the similarity between the collected data distribution and target distribution, and introduces an algorithm combining vehicle mobility prediction and ride request prediction to obtain an approximate-optimal vehicle dispatching strategy in an efficient way. The preliminary experiment based on the real-world data shows a significant improvement of the sensing coverage.

The leapfrog integrator is routinely used within the Hamiltonian Monte Carlo method and its variants. We give strong numerical evidence that alternative, easy to implement algorithms yield fewer rejections with a given computational effort. When the dimensionality of the target distribution is high, the number of accepted proposals may be multiplied by a factor of three or more. This increase in the number of accepted proposals is not achieved by impairing any positive features of the sampling. We also establish new non-asymptotic and asymptotic results on the monotonic relationship between the expected acceptance rate and the expected energy error. These results further validate the derivation of one of the integrators we consider and are of independent interest.

Adaptive Markov Chain Monte Carlo (MCMC) algorithms attempt to `learn' from the results of past iterations so the Markov chain can converge quicker. Unfortunately, adaptive MCMC algorithms are no longer Markovian, so their convergence is difficult to guarantee. In this paper, we develop new diagnostics to determine whether the adaption is still improving the convergence. We present an algorithm which automatically stops adapting once it determines further adaption will not increase the convergence speed. Our algorithm allows the computer to tune a `good' Markov chain through multiple phases of adaption, and then run conventional non-adaptive MCMC. In this way, the efficiency gains of adaptive MCMC can be obtained while still ensuring convergence to the target distribution.

У статті розглядається підхід до побудови модуля цілерозподілу в складі автоматизованої системи управління Сухопутних військ Збройних Сил України. Для отримання близького до оптимального цілерозподілу запропоновано використати генетичні алгоритми.

As a means of improving the accuracy of Raman spectroscopic quantitative analysis, a strategy combining the generation of a series of temperature-varied spectra to make diverse and characteristic structural information of sample components widely available for calibration, and subsequent selection of more property-descriptive variables among these spectra, has been demonstrated. For the evaluation, Raman spectra of synthetic hydrocarbon mixtures, lube base oils (LBOs) and polyethylene (PE) pellets were acquired at regular intervals while the sample temperature gradually increased from cryogenic to near room temperature. To select target-descriptive variables from all of the snapshot (temperature-varied) spectra, a Markov blanket (MB) feature (variable) selection able to produce a minimal set of features without changing the original target distribution was adopted. The selection utilizes a conditional independence test to quickly obtain an optimal feature subset by simultaneously considering relev...

We observe that several existing policy gradient methods (such as vanilla policy gradient, PPO, A2C) may suffer from overly large gradients when the current policy is close to deterministic (even in some very simple environments), leading to an unstable training process. To address this issue, we propose a new method, called \emph{target distribution learning} (TDL), for policy improvement in reinforcement learning. TDL alternates between proposing a target distribution and training the policy network to approach the target distribution. TDL is more effective in constraining the KL divergence between updated policies, and hence leads to more stable policy improvements over iterations. Our experiments show that TDL algorithms perform comparably to (or better than) state-of-the-art algorithms for most continuous control tasks in the MuJoCo environment while being more stable in training., Comment: AAAI-20 (oral)

Optimising the aiming strategy is crucial for Solar Power Tower plants, in order to maximise the energy generated, whilst also preventing catastrophic damage to receiver components. In this work, a bi-objective optimisation model is developed to find optimal aiming strategies for a Solar Power Tower plant. The primary objective to maximise the radiation captured by the receiver is offset by a secondary objective to minimise the deviation from a desired target distribution, which is designed by solar plant operators to improve plant efficiency. A numerical method is proposed to solve the optimisation problem, and an illustrative example is presented to show functionality of the model and the numerical method. Conclusions are drawn on the model presented, extensions are considered and current work is discussed.

The No-U-Turn Sampler (NUTS) is a relatively new Markov chain Monte Carlo (MCMC) algorithm that avoids the random walk behavior that common MCMC algorithms such as Gibbs sampling or Metropolis Hastings usually exhibit. Given the fact that NUTS can efficiently explore the entire space of the target distribution, the sampler converges to high-dimensional target distributions more quickly than other MCMC algorithms and is hence less computational expensive. The focus of this study is on applying NUTS to one of the complex IRT models, specifically the two-parameter mixture IRT (Mix2PL) model, and further to examine its performance in estimating model parameters when sample size, test length, and number of latent classes are manipulated. The results indicate that overall, NUTS performs well in recovering model parameters. However, the recovery of the class membership of individual persons is not satisfactory for the three-class conditions. Findings from this investigation provide empirical evidence on the performance of NUTS in fitting Mix2PL models and suggest that researchers and practitioners in educational and psychological measurement should benefit from using NUTS in estimating parameters of complex IRT models.

Постановка задачи. Одним из перспективных направлений развития военной авиации, которое бы значительно повысило эффективность ударных авиационных комплексов, является создание комплексов с беспилотными летательными аппаратами и развитие тактики их применения. Анализ военных конфликтов последних десятилетий показывает значительное повышение роли беспилотной авиации, как при решении задач обеспечения других видов и родов войск, так и при решении разведывательных, ударных и специальных задач. На современном этапе развития военной беспилотной авиации идут работы по созданию новых ударных беспилотных летательных аппаратов, способных решать задачи в составе группы. В связи с этим актуально решение задачи по обоснованию структуры распределённой системы управления группой ударных беспилотных летательных аппаратов и разработке алгоритмов ее функционирования, позволяющих повысить эффективность группового применения ударных беспилотных летательных аппаратов путем координации их совместных действий в автономном режиме. Целью работы является обоснование структуры распределённой системы управления группой ударных беспилотных летательных аппаратов и разработка алгоритмов координации их взаимодействия для максимизации целевого показателя на примере атаки нескольких не одинаково важных целей в автономном режиме. Используемые методы. При обосновании структуры распределенной системы управления группой беспилотных летательных аппаратов и разработке алгоритма координации их взаимодействия использовались методы системного анализа. Для синтеза показателя эффективности координированных действий беспилотных летательных аппаратов использовались метод анализа иерархий и методы линейного программирования, а для оценки эффективности алгоритма координации действий беспилотных летательных аппаратов в различных ситуациях – метод имитационного моделирования. Новизна. Элементами новизны является развитие теории функционирования системы управления группой беспилотных летательных аппаратов, позволяющей организовывать их координированные действия в автономном режиме на основе теории мультиагентных систем, а также разработан новый интегральный показатель качества оценки групповых действий беспилотных летательных аппаратов. Результат. Использование представленной распределённой сетецентрической системы управления группой беспилотных летательных аппаратов позволяет повысить эффективность групповой атаки ударными беспилотными летательными аппаратами до 23%, средний выигрыш составил 16%. Практическая значимость. Предлагаемая авторами распределенная сетецентрическая система управления группой ударных беспилотных летательных аппаратов и алгоритмы ее функционирования могут быть использованы разработчиками беспилотных летательных аппаратов при проектировании новых образцов авиационной техники., Problem statement. Promising areas of the development of military aviation, which would significantly increase the effectiveness of attack aircraft complexes, are unmanned aerial vehicles complexes creating and the development of tactics for their use. Recent decades military conflicts analysis shows a significant increase of the role of unmanned aviation, both in solving the tasks of providing other troops types, and in solving intelligence, attack and special tasks. At the present stage of the development of military unmanned aviation, activities are underway to create new strike unmanned aerial vehicles capable to solve tasks as part of a group. That’s why, it is important to solve the problem of justifying the structure of the distributed control system for a group of drone unmanned aerial vehicles and developing algorithms for its operation. These algorithms make it possible to increase the efficiency of group use of drone unmanned aerial vehicles by coordinating their joint actions in autonomous mode. Purpose. Justification of the structure of the distributed control system of a group of drone unmanned aerial vehicles and the development of algorithms for coordinating their interaction to maximize the target indicator using as the example an attack of several equally important targets in the autonomous mode are aims of the paper. Methods. To justify the structure of the distributed control system for a group of unmanned aerial vehicles and to develop an algorithm for coordinating their interaction systems analysis methods were used. To synthesize the efficiency index of coordinated actions of unmanned aerial vehicles, the method of hierarchy analysis and linear programming methods were used. The method of simulation modeling was used to evaluate the effectiveness of the algorithm for coordinating actions of unmanned aerial vehicles in various situations. Novelty. The element of novelty is the development of the theory of the unmanned aerial vehicles group control system functioning, which allows to organize their coordinated actions in an autonomous mode based on the multiagent systems theory. Also a new integral indicator of the quality of assessment of group actions of unmanned aerial vehicles is developed. Result. The use of the distributed network-centric control system of a group of unmanned aerial vehicles allows to increase the effectiveness of a group attack by using the unmanned aerial vehicles attack up to 23%. The average gain is 16%. Practical relevance. The proposed distributed network-centric control system for a group of drone unmanned aerial vehicles and algorithms for its operation can be used by developers of the unmanned aerial vehicles when designing new types of aircrafts.

One main limitation of the existing optimal scaling results for Metropolis--Hastings algorithms is that the assumptions on the target distribution are unrealistic. In this paper, we consider optimal scaling of random-walk Metropolis algorithms on general target distributions in high dimensions arising from practical MCMC models from Bayesian statistics. For optimal scaling by maximizing expected squared jumping distance (ESJD), we show the asymptotically optimal acceptance rate $0.234$ can be obtained under general realistic sufficient conditions on the target distribution. The new sufficient conditions are easy to be verified and may hold for some general classes of MCMC models arising from Bayesian statistics applications, which substantially generalize the product i.i.d. condition required in most existing literature of optimal scaling. Furthermore, we show one-dimensional diffusion limits can be obtained under slightly stronger conditions, which still allow dependent coordinates of the target distribution. We also connect the new diffusion limit results to complexity bounds of Metropolis algorithms in high dimensions., Comment: 45 pages

Deubiquitinases (DUBs), a component of protein ubiquitination, regulate many biological processes and cellular pathways critical for cell survival, proliferation, genome stability and transcriptional control that are closely related to disease such as cancer. DUBs rely on their interaction with target proteins in order to function within a particular pathway. However, there is currently little information on the DUBs' target protein. To observe the target distribution and specificity of the seven DUBs family (USP, UCH, MJD, OTU, JAMM, MINDY, ZUFSP), protein partners of every member in DUBs family were extracted from IntAct, MINT and IMEx databases and protein-protein interaction (PPI) network was constructed and analysed in Cytoscape, an open source software platform for visualising molecular interaction network. The PPI network of DUBs consists of 2328 nodes and 3409 edges and follows a power law distribution that corresponds to scale-free topology. From the PPI network, DUBs interaction is divided into: (1) DUBs-ubiquitin enzymes (e.g. E2 conjugating enzyme and E3 ligase), (2) DUBs-DUBs or also known as 'DUBbing DUBs' and (3) DUBs-other target proteins. We observed that only approximately 60% of proteins were the unique targets of each DUBs family, suggesting that target preference may not be conferred according to DUBs family as in the case of ubiquitin-linkage preference. Utilising the information from this study, it is anticipated that the potential of DUBs as diseases therapeutics target could be fully explored.

Information sanitization to protect an underlying label from being inferred through a data stream is investigated in this work. The problem is posed as an optimal mapping from an underlying distribution that reveals a class/label for the data to a target distribution with minimum distortion. The optimal sanitization operation are transformed to convex optimization problems corresponding to the domain of the source and target distributions. In particular, when one of the distributions is discrete, a parallel is drawn to a biased quantization method and an efficient sub-gradient method is proposed to derive the optimal transformation. The method is extended to a real time scenario when multiple source distributions are to be mapped to a fixed target distribution without prior knowledge of the label of the streaming data, in order to defeat any hypothesis test between the labels. It is shown that even when the source label is unknown to the sanitizer, optimal distortion is possible with perfect privacy.

Traffic splitting is essential for load balancing over multiple servers, middleboxes, and paths. Often the target traffic distribution is not uniform (e.g., due to heterogeneous servers or path capacities). A natural approach is to implement traffic split in existing rule matching tables in commodity switches. In this paper we suggest an analytical study of such an approach. To do that, we relate the description of distributions in switches to signed representations of positive integers. We suggest an optimal algorithm that minimizes the number of rules needed to represent a weighted traffic distribution. Since switches often have limited rule-table space, the target distribution cannot always be exactly achieved. Accordingly, we also develop a solution that, given a restricted number of rules, finds a distribution that can be implemented within the limited space. To select among different solutions, we describe metrics for quantifying the accuracy of an approximation. We demonstrate the efficiency of the solutions through extensive experiments.