Modifying the feature-selective validation method to validate noisy data sets

This paper identifies several important aspects of current Feature Selective Validation (FSV) methodologies that are embodied in IEEE Standard 1597.1 for the Validation of CEM Computer Modeling and Simulations. The FSV method facilitates comparisons of sets of electromagnetic (EM) observable data for a given problem to determine “levels of agreement” across amplitude and feature variables. Areas of future revision to this standard are presented that will further enhance the standard's utility for performing Computational Electromagnetic (CEM) technique validation for a wide range of problems. In particular, we consider the utility of the N-dimensional FSV and revisit applications of the Amplitude Difference Measure (ADM), Feature Difference Measure (FDM) and the Global Difference Measure (GDM). This is discussed within the context of large complex system problems that present interesting challenges to the FSV method due to the potentially wide dynamic range of the data. Certain use cases for scattering cross section, system-level coupling, and large system-level EMC problems require a somewhat modified approach in computing the GDM based on how the FDM and ADM are weighted. For the current 1-D FSV, unweighted or incorrectly weighted amplitude and feature measures can potentially lead to inconclusive or even misleading results. These issues are addressed and future revisions to the IEEE Standard 1597.1 are highlighted.

Predicting the High Intensity Radiated Field (HIRF) coupling in complex system is highly required in the aeronautic industry, which can be performed by electromagnetic simulation. But the validation of electromagnetic simulation is a difficult issue which needs proper flowchart and validation methods. In this paper, the numerical simulation by the parallel Finite-Difference Time-Domain (FDTD) method is validated on a simplify model of a metallic case with apertures under the HIRF environment. Concretely, the uncertainty quantification analysis on the input parameter of incident angles is carried out for the electric field in the center of the model with the method of nonintrusive polynomial chaos (NIPC). Based on this, the feature selective validation (FSV) is performed to quantify the comparison of the experimental data and the best simulation estimate. And the results obtained from the FSV method provide a high reliability for the numerical simulation.

Feature Select Validation (FSV) is a widely used validation method for data comparison. FSV provides a quantitative standard to describe the similarity between two sets of data. In this paper, the application of the FSV technique is extended to data optimization. The raw data obtained from simulations or measurements are often non-ideal for further processing. Several techniques, such as data perturbation, can be used to improve the data quality in certain aspects. However, after modifications the new data could be very different to the original one. Using FSV as an objective function for the optimization process is discussed in this paper, in an example of causality enforcement, to ensure the enforced casual data has the minimum deviations from the original data. The results demonstrate that the proposed approach in this paper is effective for data modification and optimization.

The Feature Selective Validation (FSV) method is currently the only technique included in Standards that is specifically directed at quantifying the comparison of computational electromagnetics with a reference as part of a programme of validation. Histogram representation of the Difference Measures in FSV is a key aspect of the way in which the comparison data is represented. The current six categories (or bins) were introduced for convenience and to allow direct comparison with visual assessment as part of the verification of FSV performance. While approaches based on continuous density functions are being developed to enable FSV to act as an intermediate step between the original data and statistical analysis, the use of histogram-based representation remains a useful and computationally simple approach. This paper investigates the optimization of the histograms used in FSV.

It has been found that the feature-selective validation (FSV) method may demonstrate inconsistencies in the results when applied to the comparison of zero-crossing datasets. This type of data is typical of time-domain-based electromagnetic compatibility validation. This paper investigates the source of this inconsistency and proposes a solution. The reason was investigated using a set of typical transient data, which was related to the derivation of the formulas of FSV. It is demonstrated that the problem can be alleviated by enhancing the definition of the offset difference measure. The resulting enhanced performance of FSV is assessed by comparing the results with visual assessment. It is demonstrated that the improvement increases the agreement between FSV prediction and visual assessment. Meanwhile, this modification has a very limited effect on the comparison of other data structures which do not cross the axis.

This paper is concerned with the usability of the feature selective validation (FSV) method for comparing validation data. It addresses how it can help in the validation process and how the resulting validation data can be interpreted. It does this by considering two different case studies. The first considers two approaches to modeling electric fields in an equipment rack, and the second compares a Spice model for coupled circuit boards with a reference full wave model. The paper concludes that while the single value summary metrics are helpful to give an overall level of agreement, the detailed point-by-point information is very helpful when considering how to improve the models or measurements involved.

The increasing demand for Power-over-Ethernet (PoE) applications in Internet of Things (IoT) infrastructure has led to a rise in the use of augmented cables as it is cost-effective. Augmented category 6 (Cat 6a) cable is now the standard for new installations requiring the aforesaid functions. Cat 6a cables for POE applications are required to be able to withstand the effects of repeated coiling they would be subjected to during installation. There is a paucity of literature on the effect of coiling on the impedance profiles of augmented cables across their length. The drawback of the available pieces of literature is that the effect of coiling is based on the frequency of operation. Examining the effect of coiling on the impedance profiles of Cat 6a cables across their length can help determine their physical integrity and aid fault location. Therefore, a method that can be used to examine the effect of coiling on impedance profiles of augmented cables across their length using the Feature Selective Validation (FSV) technique is provided. The ability of the FSV to accurately present the comparison between two data sets as a comprehensible output makes it better than other common methods. Three Cat 6a cables from different manufacturers were selected for the experiment. The Cat 6a cables were exposed to two rounds of coiling and uncoiling to imitate the stress anticipated from handling during installation. The FSV results revealed the cables with the lowest and highest impedance profile variations from the stress tests. The approach presented showed that it can be used to undertake an objective quantification of the effect of coiling on the impedance profiles of the cables across the length.

There are a number of reasons why a numerical value is a desirable outcome for the validation of numerical models: these include the desire to report objectively and succinctly on the comparison with a benchmark result, and the need to rank order implementations of a canonical problem. The key challenge is to develop techniques which provide a numerical output that do not challenge excessively the interpretations of experienced engineers and which presents both sufficient discrimination between comparisons and consistency across the various sub-disciplines of EMC. While this paper concentrates on the subject of numerical code generation, the rationale and associated concepts are just as relevant to other aspects of EMC such as quantifying experimental repeatability. The purpose of this paper is to present some of the work done to date on generating a numerical value representing the overall quality of comparison for two sets of data. This paper reviews and contrasts various techniques currently available to produce this numerical overview. The techniques include correlograms, feature selective validation (FSV), integrated error against log frequency (IELF) and reliability factors. The paper concludes that the most taxing aspect of the work required to develop such techniques is benchmarking against human interpretation, a rating scale is discussed which can assist in obtaining this information.

The use of Ethernet cables is a vital, if under- discussed element of the infrastructure for the internet of things (IOT). While there are many cable types on the market, one worrying trend is the wide availability of copper clad aluminum (CCA) cables, which are widely considered unsuitable for infrastructure deployment. The availability of these copper clad aluminum (CCA) cables frequently disguised as compliant Ethernet communication cables calls for a method of assessing their performance, as this is crucial to ensuring quality of service delivery. This paper presents a method of analyzing the measured return loss and impedance profile due to handling stress. In this research, four Ethernet cables of which one of them was copper CCA cable were subjected to three rounds of coiling and uncoiling tests to represent stress from handling during installation. The Feature Selective Validation (FSV) method and Kolmogorov-Smirnov (KS) tests were used to quantify the variations between the tests. The results indicate that the CCA cable has the lowest resilience to physical stress with high potential for degradation.

The Feature Selective Validation (FSV) method has demonstrated a number of advantages over alternative assessment methods for validation of computational electromagnetics. For example, the availability of diagnostic information and the ability to mirror human perception. However, the FSV is a heuristically based technique and, in order to have confidence in the method itself, an understanding of how the output varies as the internal coefficients vary is important. The first, important, step in FSV is to filter the original signals into ‘low’ and ‘high’ transform portions, and this paper looks at the effect of varying the coefficients associated with this operation. It supports that the current partitioning is optimum but notes the areas where substantial care is needed in selection of the parameters.

We investigated the estimation of the ensemble channelized Hotelling observer (CHO) signal-to-noise ratio (SNR) for ordered-subset (OS) image reconstruction using noisy projection data. Previously, we computed the ensemble CHO SNR using a method for approximating the channelized covariance of OS reconstruction, which requires knowledge of the noise-free projection data. Here, we use a “plug-in” approach, in which noisy data is used in place of the noise-free data in the aforementioned channelized covariance approximation. Additionally, we evaluated the use of smoothing of the noisy projections before use in the covariance approximation. Additionally, we evaluated the use of smoothing of the noisy projections before use in the covariance calculation. The task was detection of a 10% contrast Gaussian signal within a slice of the MCAT phantom. Simulated projections of the MCAT phantom were scaled and Poisson noise was added to create 100 noisy signal-absent data sets. Simulated projections of the scaled signal were then added to the noisy background projections to create 100 noisy signal-present data set. These noisy data sets were then used to generate 100 estimates of the ensemble CHO SNR for reconstructions at various iterates. For comparison purposes, the same calculation was repeated with the noise-free data. The results, reported as plots of the average CHO SNR generated in this fashion, along with 95% confidence intervals, demonstrate that this approach works very well, and would allow optimization of imaging systems and reconstruction methods using a more accurate object model (i.e., real patient data).

Seismic data denoising is a critical component of seismic data processing, yet effectively removing erratic noise, characterized by its non-Gaussian distribution and high amplitude, remains a substantial challenge for conventional methods and deep-learning (DL) algorithms. Supervised learning frameworks typically outperform others, but they require pairs of noisy data sets alongside corresponding clean ground truth, which is impractical for real-world seismic data sets. In contrast, unsupervised learning (UL) methods, which do not rely on ground truth during training, often fall short in performance when compared with their supervised or traditional denoising counterparts. Moreover, current unsupervised DL methods fail to address the specific challenges posed by erratic seismic noise adequately. This paper introduces a novel zero-shot unsupervised DL framework designed specifically to mitigate random and erratic noise, with a particular emphasis on blended noise. Drawing inspiration from Noise2Noise (N2N) and data augmentation principles, we develop a robust self-supervised denoising network called robust Noiser2Noiser. Our approach eliminates the need for paired noisy and clean data sets as required by supervised methods or paired noisy data sets as in N2N. Instead, our framework relies solely on the original noisy seismic data set. Our methodology generates two independent recorrupted data sets from the original noisy data set, using one as the input and the other as the training target. Subsequently, we use a DL-based denoiser, denoising convolutional neural network, for training purposes. To address various types of random and erratic noise, the original noisy data set is recorrupted with the same noise type. Detailed explanations for generating training input and target data for blended data are provided. We apply our network to synthetic and real marine data examples, demonstrating significantly improved noise attenuation performance compared with traditional denoising methods and state-of-the-art UL Codes are available on https://github.com/Ji-seismic/N2N_deblending.

<div> <p>Geo-Wiki is an online platform for involving citizens in the visual interpretation of very high-resolution satellite imagery to collect reference data on land cover and land use. Instead of being an ongoing citizen science project, short intensive campaigns are organized in which citizens participate. The advantage of this approach is that large amounts of data are collected in a short amount of time with a clearly defined data collection target to reach. Participants can also schedule their time accordingly, with their past feedback indicating that this intensive approach was preferred. The reference data are then used in further scientific research to answer a range of questions such as: How much of the land’s surface is wild or impacted by humans?  What is the size of agricultural fields globally? The campaigns are organized as competitions with prizes that include Amazon vouchers and co-authorship on a scientific publication. The scientific publication is the mechanism by which the data are openly shared so that other researchers can use this reference data set in other applications. The publication is usually in the form of a data paper, which explains the campaign in detail along with the data set collected. The data are uploaded to a repository such as Pangaea, ZENODO or IIASA’s own data repository, DARE.  This approach from data collection, to opening up the data, to documentation via a scientific data paper also ensures transparency in the data collection process. There have been several Geo-Wiki citizen science campaigns that have been run over the last decade. Here we provide examples of experiences from five recent campaigns: (i) the Global Cropland mapping campaign to build a cropland validation data set; (ii) the Global Field Size campaign to characterize the size of agricultural fields around the world; (iii) the Human Impact on Forests campaign to produce the first global map of forest management; (iv) the Global Built-up Surface Validation campaign to collect data on built-up surfaces for validation of global built-up products such as the Global Human Settlement Layer (https://ghsl.jrc.ec.europa.eu/); and (v) the Drivers of Tropical Forest Loss campaign, which collected data on the main causes of deforestation in the tropics. In addition to outlining the campaign, the data sets collected and the sharing of the data online, we provide lessons learned from these campaigns, which have built upon experiences collected over the last decade. These include insights related to the quality and consistency of the classifications of the volunteers including different volunteer behaviors; best practices in creating control points for use in the gamification and quality assurance of the campaigns; different methods for training the volunteers in visual interpretation; difficulties in the interpretation of some features, which may need expert input instead as well as the inability of some features to be recognized from satellite imagery; and limitations in the approach regarding change detection due to temporal availability of open satellite imagery, among several others. </p> </div>

This paper based on four key abilities of the Prognostics and Health Management (PHM) System including the abilities of fault detection, fault isolation, fault prognosis and prognosis of the remaining life, analyzes and constructs a framework of the PHM quantitative requirements, and defines the quantitative assessment emphasis of the PHM abilities. Targeting these quantitative requirements and proceeding from the viewpoint of life cycle assessment, this paper presents three methods of the PHM System validation, which are the validation method based on analysis and assessment, the validation method based on simulation (including entire simulation validation and half-entity simulation validation) and the validation method based on experimentation. This paper defines the emphasis on the PHM System analysis and assessment, ascertains the implementary elements and the simulated environmental framework of the PHM System simulated validation and elaborates two central elements of the PHM System experimental validation - the experiment scheme and the experiment technique. This paper provides thoughts and methods to the quantitative validation assessment of the rapidly developing and applying PHM System.

The aim of this article is to present a dynamic record linkage solution using high level machine learning to solve the problems associated with noisy, inconsistent and dynamic data sets. Classic deterministic and probabilistic models fail to work in such environments because they are static and based on assumptions. Incorporating supervised learning algorithm, active learning, and ensemble techniques such as random forests and boosting allows the proposed structure to change in response to different data types and thus become more accurate and scaleable. Some of its highlights are feature selection to ensure match accuracy, clustering to support noisy inputs, and active learning to minimize reliance on large labeled datasets. Simulations with real data, such as government or healthcare data, show that compared with the traditional approach, linkage is significantly more accurate and efficient. The flexible model led to 15% higher F1-scores on noisy datasets and was scalable across large data sets. This work demonstrates how adaptive machine learning can revolutionize the modern record linkage tasks and provides a powerful and effective solution to ever-changing data conditions.