Estimation Of Distribution Research Articles

DW-GAN: Toward High-Fidelity Color-Tones of GAN-Generated Images With Dynamic Weights.

Color-tone represents the prominent color of an image, and training generative adversarial nets (GAN) to change color-tones of generated images is desirable in many applications. Advances such as HistoGAN can manipulate color-tones of generated images with a target image. Yet, there are challenges. Kullback-Leibler (KL) divergence adopted by HistoGAN might bring the color-tone mismatching, because it is possible to provide infinite score to a generator. Moreover, only relying on distribution estimation also produces images with lower fidelity in HistoGAN. To address these issues, we propose a new approach, named dynamic weights GAN (DW-GAN). We use two discriminators to estimate the distribution matching degree and details' similarity, with Laplacian operator and Hinge loss. Laplacian operator can help capture more image details, while Hinge loss is deduced from mean difference (MD) that could avoid the case of infinite score. To synthesize desired images, we combine the loss of the two discriminators with generator loss and set the weights of the two estimated scores to be dynamic through the previous discriminators' outputs, given that the training signal of a generator is from a discriminator. Besides, we innovatively integrate the dynamic weights into other GAN variants (e.g., HistoGAN and StyleGAN) to show the improved performance. Finally, we conduct extensive experiments on one industrial Fabric and seven public datasets to demonstrate the significant performance of DW-GAN in producing higher fidelity images and achieving the lowest Frechet inception distance (FID) scores over SOTA baselines.

Feature Separation in Diffuse Lung Disease Image Classification by Using Evolutionary Algorithm-Based NAS.

In the field of diagnosing lung diseases, the application of neural networks (NNs) in image classification exhibits significant potential. However, NNs are considered "black boxes," making it difficult to discern their decision-making processes, thereby leading to skepticism and concern regarding NNs. This compromises model reliability and hampers intelligent medicine's development. To tackle this issue, we introduce the Evolutionary Neural Architecture Search (EvoNAS). In image classification tasks, EvoNAS initially utilizes an Evolutionary Algorithm to explore various Convolutional Neural Networks, ultimately yielding an optimized network that excels at separating between redundant texture features and the most discriminative ones. Retaining the most discriminative features improves classification accuracy, particularly in distinguishing similar features. This approach illuminates the intrinsic mechanics of classification, thereby enhancing the accuracy of the results. Subsequently, we incorporate a Differential Evolution algorithm based on distribution estimation, significantly enhancing search efficiency. Employing visualization techniques, we demonstrate the effectiveness of EvoNAS, endowing the model with interpretability. Finally, we conduct experiments on the diffuse lung disease texture dataset using EvoNAS. Compared to the original network, the classification accuracy increases by 0.56%. Moreover, our EvoNAS approach demonstrates significant advantages over existing methods in the same dataset.

Bayesian estimation of 3-component mixture geometric distribution under different loss functions

This article focuses on discrete survival data analysis, employing a doubly Type-I censoring scheme (DT1SC) under a Bayesian framework for parameters estimation in a 3-component mixture geometric (3-CMG) distribution. The non-informative (uniform) prior is utilized under squared error, DeGroot, and precautionary loss functions. The time is considered as discrete in this paper, presenting a departure from continuous approaches in survival analysis. The methodology is tailored to address challenges that traditional survival analysis encounters when faced with limited or missing information. The proposed method effectively handles the complexities arising from incomplete or missing data, providing a robust mechanism for accurate parameter estimation. Through extensive simulations and application to real-world datasets, our approach demonstrates its effectiveness in managing uncertainties, particularly in scenarios involving sparse or missing data. This work represents a noteworthy contribution to methodological advancements in survival analysis, offering a valuable tool for researchers and practitioners navigating intricate dynamics within the 3-CMG under DT1CS with time being considered as discrete.

The multi-parameter estimation of discrete distribution without closed-form solutions by the US algorithm

<p>Strong and steady convergence characterizes the upper-crossing/solution (US) algorithm, which is an effective method for identifying roots of a complicated nonlinear equation $ h(\theta) = 0 $. Only the case where one parameter of a distribution function can be directly specified by another parameter is taken into account by the research that is currently available. However, whether this approach can be applied in multi-parameter scenarios where one parameter cannot be clearly represented by the other is an issue deserving of more investigation. In order to extend the applicability of the US algorithm, this article used the Type Ⅰ discrete Weibull distribution with two parameters as an example. It then combined the US algorithm with the first-derivative lower bound (FLB) function method to estimate the complex situation where two parameters cannot be expressed as each other. Simulation studies and empirical analysis demonstrated that the US algorithm performs more accurately and steadily than the traditional Newton method.</p>

A Novel Poroelastography Method for High-quality Estimation of Lateral Strain, Solid Stress and Fluid Pressure In Vivo.

Assessment of mechanical and transport properties of tissues using ultrasound elasticity imaging requires accurate estimations of the spatiotemporal distribution of volumetric strain. Due to physical constraints such as pitch limitation and the lack of phase information in the lateral direction, the quality of lateral strain estimation is typically significantly lower than the quality of axial strain estimation. In this paper, a novel lateral strain estimation technique based on the physics of compressible porous media is developed, tested and validated. This technique is referred to as "Poroelastography-based Ultrasound Lateral Strain Estimation" (PULSE). PULSE differs from previously proposed lateral strain estimators as it uses the underlying physics of internal fluid flow within a local region of the tissue as theoretical foundation. PULSE establishes a relation between spatiotemporal changes in the axial strains and corresponding spatiotemporal changes in the lateral strains, effectively allowing assessment of lateral strains with comparable quality of axial strain estimators. We demonstrate that PULSE can also be used to accurately track compression-induced solid stresses and fluid pressure in cancers using ultrasound poroelastography (USPE). In this study, we report the theoretical formulation for PULSE and validation using finite element (FE) and ultrasound simulations. PULSE-generated results exhibit less than 5% percentage relative error (PRE) and greater than 90% structural similarity index (SSIM) compared to ground truth simulations. Experimental results are included to qualitatively assess the performance of PULSE in vivo. The proposed method can be used to overcome the inherent limitations of non-axial strain imaging and improve clinical translatability of USPE.

A joint model for the estimation of species distributions and environmental characteristics from point-referenced data.

Predicting and explaining species occurrence using environmental characteristics is essential for nature conservation and management. Species distribution models consider species occurrence as the dependent variable and environmental conditions as the independent variables. Suitable conditions are estimated based on a sample of species observations, where one assumes that the underlying environmental conditions are known. This is not always the case, as environmental variables at broad spatial scales are regularly extrapolated from point-referenced data. However, treating the predicted environmental conditions as accurate surveys of independent variables at a specific point does not take into account their uncertainty. We present a joint hierarchical Bayesian model where models for the environmental variables, rather than a set of predicted values, are input to the species distribution model. All models are fitted together based only on point-referenced observations, which results in a correct propagation of uncertainty. We use 50 plant species representative of the Dutch flora in natural areas with 8 soil condition predictors taken during field visits in the Netherlands as a case study. We compare the proposed model to the standard approach by studying the difference in associations, predicted maps, and cross-validated accuracy. We find that there are differences between the two approaches in the estimated association between soil conditions and species occurrence (correlation 0.64-0.84), but the predicted maps are quite similar (correlation 0.82-1.00). The differences are more pronounced in the rarer species. The cross-validated accuracy is substantially better for 5 species out of the 50, and the species can also help to predict the soil characteristics. The estimated associations tend to have a smaller magnitude with more certainty. These findings suggests that the standard model is often sufficient for prediction, but effort should be taken to develop models which take the uncertainty in the independent variables into account for interpretation.

Tumor-treating fields dosimetry in glioblastoma: Insights into treatment planning, optimization, and dose-response relationships.

Tumor-treating fields (TTFields) are currently a Category 1A treatment recommendation by the US National Comprehensive Cancer Center for patients with newly diagnosed glioblastoma. Although the mechanism of action of TTFields has been partly elucidated, tangible and standardized metrics are lacking to assess antitumor dose and effects of the treatment. This paper outlines and evaluates the current standards and methodologies in the estimation of the TTFields distribution and dose measurement in the brain and highlights the most important principles governing TTFields dosimetry. The focus is on clinical utility to facilitate a practical understanding of these principles and how they can be used to guide treatment. The current evidence for a correlation between TTFields dose, tumor growth, and clinical outcome will be presented and discussed. Furthermore, we will provide perspectives and updated insights into the planning and optimization of TTFields therapy for glioblastoma by reviewing how the dose and thermal effects of TTFields are affected by factors such as tumor location and morphology, peritumoral edema, electrode array position, treatment duration (compliance), array "edge effect," electrical duty cycle, and skull-remodeling surgery. Finally, perspectives are provided on how to optimize the efficacy of future TTFields therapy.

On approach for prognosis of dissolution of a medicinal product in an organism

In this article, we introduce a model of dissolution of a medicinal product in an organism with account of changing of conditions. The model based on estimation of spatiotemporal distribution of concentration of the above product during dissolution. We introduce an analytical approach for analysis of the considered dissolution. We consider a possibility to accelerate and decelerate of the above dissolution.

Score-Based Counterfactual Generation for Interpretable Medical Image Classification and Lesion Localization.

Deep neural networks (DNNs) have immense potential for precise clinical decision-making in the field of biomedical imaging. However, accessing high-quality data is crucial for ensuring the high-performance of DNNs. Obtaining medical imaging data is often challenging in terms of both quantity and quality. To address these issues, we propose a score-based counterfactual generation (SCG) framework to create counterfactual images from latent space, to compensate for scarcity and imbalance of data. In addition, some uncertainties in external physical factors may introduce unnatural features and further affect the estimation of the true data distribution. Therefore, we integrated a learnable FuzzyBlock into the classifier of the proposed framework to manage these uncertainties. The proposed SCG framework can be applied to both classification and lesion localization tasks. The experimental results revealed a remarkable performance boost in classification tasks, achieving an average performance enhancement of 3-5% compared to previous state-of-the-art (SOTA) methods in interpretable lesion localization.

Random permutations generated by delay models and estimation of delay distributions

Random permutations generated by delay models and estimation of delay distributions

An observer for solar concentrator fields based on multiple measurements

The estimation of the temperature distribution in solar concentration plants is crucial for monitoring and control purposes. These plants cover large surfaces, and therefore, partial shading can occur. Under such conditions, solar irradiance varies according to the location of the collectors. This work addresses the problem of designing an observer to estimate the temperature profile of several parabolic solar concentrators based on multiple measurements. The convergence of the proposed observer is analyzed using Lyapunov theory, and its performance is illustrated through simulations. These results demonstrate that the use of additional measurements improves the observer's convergence under partial shading. Further work is proposed to address tuning issues and the effects of uncertain irradiance measurements.

CORE: CORrelation-Guided Feature Enhancement for Few-Shot Image Classification.

Few-shot classification aims to adapt classifiers trained on base classes to novel classes with a few shots. However, the limited amount of training data is often inadequate to represent the intraclass variations in novel classes. This can result in biased estimation of the feature distribution, which in turn results in inaccurate decision boundaries, especially when the support data are outliers. To address this issue, we propose a feature enhancement method called CORrelation-guided feature Enrichment that generates improved features for novel classes using weak supervision from the base classes. The proposed CORrelation-guided feature Enhancement (CORE) method utilizes an autoencoder (AE) architecture but incorporates classification information into its latent space. This design allows the CORE to generate more discriminative features while discarding irrelevant content information. After being trained on base classes, CORE's generative ability can be transferred to novel classes that are similar to those in the base classes. By using these generative features, we can reduce the estimation bias of the class distribution, which makes few-shot learning (FSL) less sensitive to the selection of support data. Our method is generic and flexible and can be used with any feature extractor and classifier. It can be easily integrated into existing FSL approaches. Experiments with different backbones and classifiers show that our proposed method consistently outperforms existing methods on various widely used benchmarks.

Beam diagnostics with silicon pixel detector array at PADME experiment

During 2022 data taking (Run III) PADME searched for a resonant production and a visible decay of the X17 particle into e+e-. A precise knowledge within 1% uncertainty of the number of positrons was required for the observation. To that purpose, an array of 2 × 6 Timepix3 (total of 512 × 1536 pixels) hybrid pixel detectors operated in data-streaming mode with ToA resolution of 1.56 ns for every pixel was employed. Two methods for data acquisition were developed. A frame-based method, integrating the number of hits for each individual pixel for a predefined period of time served for monitoring the beam conditions and to provide a rough estimation of the beam distribution and number of positrons. A data streaming mode exploiting the nanosecond time resolution of Timepix3 detector was used for precise characterization of the transverse beam profile and the distribution of the incident positrons within each bunch of ∼ 200 ns duration.

Optimizing Weibull Distribution Parameters for Improved Earthquake Modeling in Japan: A Comparative Approach

The Weibull distribution is considered one of the important distributions used in reliability and in the distribution of survival times and in neutrosophic prediction. This paper contained an estimate of a two-parameter Weibull distribution using three estimation methods. One of these methods for estimating parameters is the traditional method, which represents the estimation of the greatest possibility, and the other two methods are estimation using the sine algorithm. and cosine (SCA) and the Ant Colony Algorithm (ACO). The simulation method was used to compare the methods, and it was found that the best method for estimating the parameters of the Weibull distribution is the sine and cosine algorithm (SCA) method. Then, real data was used, represented by the intensity of the earthquake in Japan (Richter) for the period July 1, 2023, to July 16, 2023, to estimate the rate of earthquake intensity in Japan, since Japan is one of the countries most exposed to earthquakes, and it was shown from The results are that the average magnitude of the earthquake in Japan in the period studied is 3.328593, which can be said. Weak buildings may be greatly damaged, but strong buildings are not greatly damaged. Also, a neutrosophic simulation of the same set of data will be suggested for future applications.

Estimating the potential risk of transmission of arboviruses in the Americas and Europe: a modelling study

Estimates of the spatiotemporal distribution of different mosquito vector species and the associated risk of transmission of arboviruses are key to design adequate policies for preventing local outbreaks and reducing the number of human infections in endemic areas. In this study, we quantified the abundance of Aedes albopictus and Aedes aegypti and the local transmission potential for three arboviral infections at an unprecedented spatiotemporal resolution in areas where no entomological surveillance is available. We developed a computational model to quantify the daily abundance of Aedes mosquitoes, leveraging temperature and precipitation records. The model was calibrated on mosquito surveillance data collected in 115 locations in Europe and the Americas between 2007 and 2018. Model estimates were used to quantify the reproduction number of dengue virus, Zika virus, and chikungunya in Europe and the Americas, at a high spatial resolution. In areas colonised by both Aedes species, A aegypti was estimated to be the main vector for the transmission of dengue virus, Zika virus, and chikungunya, being associated with a higher estimate of R0 when compared with A albopictus. Our estimates highlighted that these arboviruses were endemic in tropical and subtropical countries, with the highest risks of transmission found in central America, Venezuela, Colombia, and central-east Brazil. A non-negligible potential risk of transmission was also estimated for Florida, Texas, and Arizona (USA). The broader ecological niche of A albopictus could contribute to the emergence of chikungunya outbreaks and clusters of dengue autochthonous cases in temperate areas of the Americas, as well as in mediterranean Europe (in particular, in Italy, southern France, and Spain). Our results provide a comprehensive overview of the transmission potential of arboviral diseases in Europe and the Americas, highlighting areas where surveillance and mosquito control capacities should be prioritised. EU and Ministero dell'Università e della Ricerca, Italy (Piano Nazionale di Ripresa e Resilienza Extended Partnership initiative on Emerging Infectious Diseases); EU (Horizon 2020); Ministero dell'Università e della Ricerca, Italy (Progetti di ricerca di Rilevante Interesse Nazionale programme); Brazilian National Council of Science, Technology and Innovation; Ministry of Health, Brazil; and Foundation of Research for Minas Gerais, Brazil.

Empirically adjusted fixed-effects meta-analysis methods in genomic studies.

In recent years, meta-analyzing summary results from multiple studies has become a common practice in genomic research, leading to a significant improvement in the power of statistical detection compared to an individual genomic study. Meta analysis methods that combine statistical estimates across studies are known to be statistically more powerful than those combining statistical significance measures. An approach combining effect size estimates based on a fixed-effects model, called METAL, has gained extreme popularity to perform the former type of meta-analysis. In this article, we discuss the limitations of METAL due to its dependence on the theoretical null distribution, leading to incorrect significance testing results. Through various simulation studies and real genomic data application, we show how modifying the z-scores in METAL, using an empirical null distribution, can significantly improve the results, especially in presence of hidden confounders. For the estimation of the null distribution, we consider two different approaches, and we highlight the scenarios when one null estimation approach outperforms the other. This article will allow researchers to gain an insight into the importance of using an empirical null distribution in the fixed-effects meta-analysis as well as in choosing the appropriate empirical null distribution estimation approach.

Analysis of Weibull progressively first-failure censored data with beta-binomial removals

<p>This study examined the estimations of Weibull distribution using progressively first-failure censored data, under the assumption that removals follow the beta-binomial distribution. Classical and Bayesian approaches for estimating unknown model parameters have been established. The estimations included scale and shape parameters, reliability and failure rate metrics as well as beta-binomial parameters. Estimations were considered from both point and interval viewpoints. The Bayes estimates were developed by using the squared error loss and generating samples for the posterior distribution through the Markov Chain Monte Carlo technique. Two interval estimation approaches are considered: approximate confidence intervals based on asymptotic normality of likelihood estimates and Bayes credible intervals. To investigate the performance of classical and Bayesian estimations, a simulation study was considered by various kinds of experimental settings. Furthermore, two examples related to real datasets were thoroughly investigated to verify the practical importance of the suggested methodologies.</p>

Comparing acoustic and satellite telemetry: an analysis quantifying the space use of Chelonia mydas in Bimini, Bahamas.

Passive acoustic and Argos satellite telemetry are common methods for tracking marine species and are often used similarly to quantify space use. However, data-driven comparisons of these methods and their associated ecological inferences are limited. To address this, we compared temporal durations, spatial resolutions, financial costs and estimates of occurrence and range distributions for each tracking approach using nine juvenile green turtles (Chelonia mydas) in Bimini, Bahamas. Tracking durations were similar, although acoustic tracking provided higher spatiotemporal resolution than satellite tracking. Occurrence distributions (95%) estimated from satellite telemetry were 12 times larger than those from acoustic telemetry, while satellite range distributions (95%) were 89 times larger. While individuals generally remained within the extent of the acoustic receiver array, gaps in coverage were identified. These gaps, combined with the lower accuracy of satellite telemetry, were likely drivers for the larger satellite distributions. Costs differed between telemetry methods, with acoustic telemetry being less expensive at larger sample sizes with a previously established array. Our results suggest that acoustic and satellite telemetry may not provide similar inferences of individual space use. As such, we provide recommendations to identify telemetry methods appropriate for specific study objectives and provide discussion on the biases of each.

Estimation of neutron fluence distribution within German PWR components for decommissioning studies

On 15 April 2023, the last German NPPs were shut down. The final shutdown is followed by a post-operational phase in which measures can be carried out to prepare for the NPPs dismantling and decommissioning. One of the tasks in preparation for the dismantling is to acquire accurate knowledge of the activation distribution within the reactor components. The activation levels depend on the neutron fluence within the reactor. Therefore, to achieve this task, a 3D detailed Monte Carlo model of a German PWR has been developed to calculate the neutron fluence characteristics (spectrum and distribution) within the components of such nuclear reactors type. The reliability of the calculations was validated via metal foil-activation measurements carried out in two active PWRs. This paper presents the Monte Carlo model and describes the calculations and measurement procedures.

Exploring English Translation Strategies Oriented by Big Data Technology

Abstract English translation is not a simple task, and it needs to be translated on the basis of different languages and cultures in order to get the ideal translation effect and promote cross-cultural communication between different languages. In this paper, with the help of web crawler technology to crawl the Chinese-English translation corpus data, and use the inverse document frequency and dynamic planning ideas to align the corpus data, the Chinese-English translation corpus is established. The LDA model is used to mine subject information in the Chinese-English translation corpus, and edge distribution estimation is used to construct the English translation model. For the application of the translation model and the Chinese-English translation corpus in improving the quality and effect of English translation, the word and sentence features, target word blocks, and the comparison of language features of human-computer translation from the Chinese-English corpus are verified in various aspects. The results show that the class/formant ratio of the source language in the corpus is between 56.73% and 75.84%, its average word length fluctuates in the range of [3.91,5.62], and the target word chunks of the English text are mainly based on nominal and verbal structures. The English translation model has a less significant discrepancy in other language features than human translation, except for the difference of 4.526 between sentence lengths. English translation needs to adopt a reasonable annotation method, semantic conversion, and technical tools to improve the quality of English translation.