Data Perturbation Research Articles

Impact localization and force reconstruction for composite plates based on virtual time reversal processing with time-domain spectral finite element method

Composite structures are widely used in different industries as important load-carrying components due to their excellent performance. These structures are prone to be damaged under low-velocity impacts with invisible cracks. Hence, impact identification is a critical part of composite structural health monitoring. In this paper, an approach is developed, which is based on time-reversal (T-R) processing and time-domain spectral finite element method. The spatial and temporal focusing properties of T-R are utilized to locate the impact source and restore the impact force, respectively. Moreover, the impact energy is evaluated by establishing the function of impact force and kinetic energy at the impact moment. The proposed method is validated experimentally for different scenarios. The influences of the signal length and accuracy of the numerical model on the proposed method are discussed in detail. The results demonstrate that the proposed method can correctly locate the impact source in composite structures, and the identified impact force can reproduce the impact event efficiently. The impact localization is essentially independent of the effect of signal length while it is vulnerable to the mesh grid. Besides, the developed method has consistent results in the case of data perturbation of material properties in the numerical model, which shows its good robustness.

A Scale-Free Approach for False Discovery Rate Control in Generalized Linear Models

ABSTRACT The Generalized Linear Model (GLM) has been widely used in practice to model counts or other types of non-Gaussian data. This article introduces a framework for feature selection in the GLM that can achieve robust False Discovery Rate (FDR) control. The main idea is to construct a mirror statistic based on data perturbation to measure the importance of each feature. FDR control is achieved by taking advantage of the mirror statistic’s property that its sampling distribution is (asymptotically) symmetric about zero for any null feature. In the moderate-dimensional setting, that is, , we construct the mirror statistic based on the maximum likelihood estimation. In the high-dimensional setting, that is, , we use the debiased Lasso to build the mirror statistic. The proposed methodology is scale-free as it only hinges on the symmetry of the mirror statistic, thus, can be more robust in finite-sample cases compared to existing methods. Both simulation results and a real data application show that the proposed methods are capable of controlling the FDR and are often more powerful than existing methods including the Benjamini-Hochberg procedure and the knockoff filter. Supplementary materials for this article are available online.

Toward Correlated Data Trading for Private Web Browsing History

The trading of social media data has attracted wide research interests over years. In particular the trading for web browsing histories, when being applied to targeted advertising, produces tremendous economic value for data consumers. However, the disclosure of entire browsing histories, even in form of anonymous datasets, poses a huge threat to user privacy. Although some existing solutions have investigated privacy-preserving outsourcing of social media data, unfortunately, they neglected the impact on the data consumer’s utility. In this paper, we propose CEATSE, a CorrElated dAta Trading framework for various kinds of private web browSing historiEs. CEATSE first models the correlation among multiple dimensional features, and then generates the optimal feature clustering scheme. Combined with this scheme, CEATSE next incorporates a correlated data perturbation strategy on each feature cluster, in order to balance the privacy-utility tradeoff. It then quantifies each chosen data contributor’s privacy loss on optimal feature clusters. Through real-data based experiments, our analysis and evaluation results demonstrate CEATSE indeed achieves user privacy protection, the data consumer’s accuracy requirement, and truthfulness, individual rationality as well as budget balance.

SCREE: a comprehensive pipeline for single-cell multi-modal CRISPR screen data processing and analysis.

Single-cell CRISPR screens have been widely used to investigate gene regulatory circuits in diverse biological systems. The recent development of single-cell CRISPR screens has enabled multimodal profiling of perturbed cells with both gene expression, chromatin accessibility and protein levels. However, current methods cannot meet the analysis requirements of different types of data and have limited functions. Here, we introduce Single-cell CRISPR screens data analysEs and perturbation modEling (SCREE) as a comprehensive and flexible pipeline to facilitate the analyses of various types of single-cell CRISPR screens data. SCREE performs read alignment, sgRNA assignment, quality control, clustering and visualization, perturbation enrichment evaluation, perturbation efficiency modeling, gene regulatory score calculation and functional analyses of perturbations for single-cell CRISPR screens with both RNA, ATAC and multimodal readout. SCREE is available at https://github.com/wanglabtongji/SCREE.

Anti-aliasing deep image classifiers using novel depth adaptive blurring and activation function

Deep convolutional networks are vulnerable to image translation or shift, partly due to common down-sampling layers, e.g., max-pooling and strided convolution. These operations violate the Nyquist sampling rate and cause aliasing. The textbook solution is low-pass filtering (blurring) before down-sampling, which can benefit deep networks as well. Even so, non-linearity units, such as ReLU, often re-introduce the problem, suggesting that blurring alone may not suffice. In this work, first, we analyse deep features with Fourier transform and show that Depth Adaptive Blurring is more effective, as opposed to monotonic blurring. To this end, we propose a novel Depth Adaptive Blur-pool (DAB-pool) module to replace existing down-sampling methods. Second, we introduce a novel activation function – with a built-in low pass filter, as an additional measure, to keep the problem from reappearing. From experiments, we observe generalisation on other forms of transformations and corruptions as well, e.g., rotation, scale, and noise. We evaluate our method under three challenging settings: (1) a variety of image translations; (2) adversarial attacks – both ℓp bounded and unbounded; and (3) data corruptions and perturbations. In each setting, our method achieves state-of-the-art results and improves clean accuracy on various benchmark datasets.

Predictive stability criteria for penalty selection in linear models

Choosing a shrinkage method can be done by selecting a penalty from a list of pre-specified penalties or by constructing a penalty based on the data. If a list of penalties for a class of linear models is given, we introduce a predictive stability criterion based on data perturbation to select a shrinkage method from the list. Simulation studies show that our predictive method identifies shrinkage methods that usually agree with existing literature and help explain heuristically when a given shrinkage method can be expected to perform well. If the preference is to construct a penalty customized for a given problem, then we propose a technique based on genetic algorithms, again using a predictive criterion. We find that, in general, a custom penalty never performs worse than any commonly used penalties and there are cases the custom penalty reduces to a recognizable penalty. Since penalty selection is mathematically equivalent to prior selection, our method also constructs priors. Our methodology allows us to observe that the oracle property typically holds for penalties that satisfy basic regularity conditions and therefore is not restrictive enough to play a direct role in penalty selection. In addition, our methodology, can be immediately applied to real data problems, and permits us to take model mis-specification into account.

Glee: A granularity filter for feature selection

In the field of Granular Computing (GrC), feature selection is an attractive task. Some basics of GrC such as information granulation and granularity have well guided the explorations of feature selection technique. Nevertheless, some challenges to perform feature selection such as high computational costs of repeating information granulation and multi-granularity structure in the process of searching qualified features may also arise. In this study, to improve the efficiency and effectiveness of feature selector, a novel framework named Glee: Granularity fiLter for fEature sElection, is reported. Firstly, the value of granularity related to each feature is calculated. Secondly, all features are reordered by the corresponding values of granularity. Finally, following such a derived sequence, features will be added into the selection pool one by one until the termination condition is achieved. Glee can not only eliminate the iterative calculations of information granulation in the whole process of selecting, but also provide sequence of features which may be insensitive to data perturbation. It should also be emphasized that Glee is a general framework, and most existing termination conditions to feature selection can be embedded into it. To validate the effectiveness of Glee, it is compared with several well-established feature selection schemes in elapsed time of selecting features, stability of selected features and classification performance. Experimental results over 20 UCI datasets with both raw features and 3 different ratios of noisy features demonstrate that our framework is superior as it yields better robustness with satisfactory elapsed time.

Stability of Riemann solutions to a class of non-strictly hyperbolic systems of conservation laws

This paper is concerned with Riemann problem for a class of non-strictly hyperbolic systems of conservation laws. The two kinds of Riemann solutions including vacuum and delta shock wave are constructed. Under the generalized Rankine-Hugoniot relation and entropy condition, we establish the existence and uniqueness of delta shock wave solutions. Furthermore, by studying the interactions among of the delta shock wave and vacuum as well as contact discontinuity, the Riemann solutions with four kinds of different structures are obtained. Additionally, the stability of the Riemann solutions is obtained under certain perturbation of the initial data.

Evaluating the Privacy and Utility of Time-Series Data Perturbation Algorithms

Data collected from sensor-rich systems may reveal user-related patterns that represent private information. Sensitive patterns from time-series data can be protected using diverse perturbation methods; however, choosing the method that provides the desired privacy and utility level is challenging. This paper proposes a new procedure for evaluating the utility and privacy of perturbation techniques and an algorithm for comparing perturbation methods. The contribution is significant for those involved in protecting time-series data collected from various sensors as the approach is sensor-type-independent, algorithm-independent, and data-independent. The analysis of the impact of data integrity attacks on the perturbed data follows the methodology. Experimental results obtained using actual data collected from a VW Passat vehicle via the OBD-II port demonstrate the applicability of the approach to measuring the utility and privacy of perturbation algorithms. Moreover, important benefits have been identified: the proposed approach measures both privacy and utility, various distortion and perturbation methods can be compared (no matter how different), and an evaluation of the impact of data integrity attacks on perturbed data is possible.

Deep Generative Mixture Model for Robust Imbalance Classification.

Discovering hidden pattern from imbalanced data is a critical issue in various real-world applications. Existing classification methods usually suffer from the limitation of data especially for minority classes, and result in unstable prediction and low performance. In this paper, a deep generative classifier is proposed to mitigate this issue via both model perturbation and data perturbation. Specially, the proposed generative classifier is derived from a deep latent variable model where two variables are involved. One variable is to capture the essential information of the original data, denoted as latent codes, which are represented by a probability distribution rather than a single fixed value. The learnt distribution aims to enforce the uncertainty of model and implement model perturbation, thus, lead to stable predictions. The other variable is a prior to latent codes so that the codes are restricted to lie on components in Gaussian Mixture Model. As a confounder affecting generative processes of data (feature/label), the latent variables are supposed to capture the discriminative latent distribution and implement data perturbation. Extensive experiments have been conducted on widely-used real imbalanced image datasets. Experimental results demonstrate the superiority of our proposed model by comparing with popular imbalanced classification baselines on imbalance classification task.

Riemann problem and wave interactions for an inhomogeneous Aw-Rascle traffic flow model with extended Chaplygin gas

The Riemann problem and wave interactions are discussed and investigated for an inhomogeneous Aw-Rascle (AR) traffic flow model with extended Chaplygin gas pressure. First, under some variable transformation, the Riemann problem with initial data of two piecewise constants is solved and two different types of Riemann solutions involving rarefaction wave, shock wave and contact discontinuity are obtained. Second, by studying the Riemann problem with three-piecewise-constant initial data, we analyze the interactions of waves and establish the global structures of Riemann solutions. It is shown that, influenced by the source term, the Riemann solutions for the inhomogeneous AR traffic flow model are no longer self-similar, and all the elementary wave curves do not keep straight. Finally, the stability of solution under the small perturbation of initial data is briefly discussed.

Point Cloud Instance Segmentation with Inaccurate Bounding-Box Annotations.

Most existing point cloud instance segmentation methods require accurate and dense point-level annotations, which are extremely laborious to collect. While incomplete and inexact supervision has been exploited to reduce labeling efforts, inaccurate supervision remains under-explored. This kind of supervision is almost inevitable in practice, especially in complex 3D point clouds, and it severely degrades the generalization performance of deep networks. To this end, we propose the first weakly supervised point cloud instance segmentation framework with inaccurate box-level labels. A novel self-distillation architecture is presented to boost the generalization ability while leveraging the cheap but noisy bounding-box annotations. Specifically, we employ consistency regularization to distill self-knowledge from data perturbation and historical predictions, which prevents the deep network from overfitting the noisy labels. Moreover, we progressively select reliable samples and correct their labels based on the historical consistency. Extensive experiments on the ScanNet-v2 dataset were used to validate the effectiveness and robustness of our method in dealing with inexact and inaccurate annotations.

Effective and Privacy-Preserving Estimation of the Density Distribution of LBS Users under Geo-Indistinguishability

With the widespread use of mobile devices, location-based services (LBSs), which provide useful services adjusted to users’ locations, have become indispensable to our daily lives. However, along with several benefits, LBSs also create problems for users because to use LBSs, users are required to disclose their sensitive location information to the service providers. Hence, several studies have focused on protecting the location privacy of individual users when using LBSs. Geo-indistinguishability (Geo-I), which is based on the well-known differential privacy, has recently emerged as a de-facto privacy definition for the protection of location data in LBSs. However, LBS providers require aggregate statistics, such as user density distribution, for the purpose of improving their service quality, and deriving them accurately from the location dataset received from users is difficult owing to the data perturbation of Geo-I. Thus, in this study, we investigated two different approaches, the expectation-maximization (EM) algorithm and the deep learning based approaches, with the aim of precisely computing the density distribution of LBS users while preserving the privacy of location datasets. The evaluation results show that the deep learning approach significantly outperforms other alternatives at all privacy protection levels. Furthermore, when a low level of privacy protection is sufficient, the approach based on the EM algorithm shows performance results similar to those of the deep learning solution. Thus, it can be used instead of a deep learning approach, particularly when training datasets are not available.

Data perturbations in stochastic generalized equations: statistical robustness in static and sample average approximated models

Data perturbations in stochastic generalized equations: statistical robustness in static and sample average approximated models

Asymptotically hyperbolic Einstein constraint equations with apparent horizon boundary and the Penrose inequality for perturbations of Schwarzschild-AdS *

We prove the existence of asymptotically hyperbolic solutions to the vacuum Einstein constraint equations with a marginally outer trapped boundary of positive mean curvature, using the constant mean curvature conformal method. As an application of this result, we verify the Penrose inequality for certain perturbations of Schwarzschild Anti-de Sitter black hole initial data.

An improved ensemble classification-based secure two stage bagging pruning technique for guaranteeing privacy preservation of DNA sequences in electronic health records

The advent of machine learning in the recent decade has excelled in determining new potential features and non-linear relationships existing between the data derived from the Electronic Health Records (EHR). Machine learning also enhances the process of handling data with maximum predictor variables compared to observations during the data mining process of prediction. The EHR data is often confronted with quality issues that are related to misclassification, missingness and measurement errors. In this context, ensemble classification schemes are determined to be essential for preventing the quality issues of EHR data. Moreover, the data sources like EHR include sensitive information that needs to be protected from disclosure before it is forwarded to the mining process. Further, the sensitive data of EHR must be hidden without modifying the dataset such that it does not influence the prediction accuracy of the incorporated ensemble classification mechanism. In this paper, the process of hiding EHR data is facilitated through Improved Sensitivity Drift based k-Anonymized Data Perturbation Scheme (ISD-k-ADP) that randomly perturbs the data in the dataset by including restricted amount of noise. This controlled amount of included noise is derived carefully from the Sensitivity Drift based depending on the expected privacy level before it is sent to the process of classification. This ISD-k-ADP scheme is reliable such that, it prevents the impact induced by the hidden data during the process of Two Stage Bagging Pruning based Ensemble Classification (TSBP-EC). Furthermore, the TSBP-EC uses the methods of distance and accuracy based pruning that aids in minimizing the size of the ensemble for ensuring effective and efficient classification using machine learning. The simulation results of the proposed ISD-k-ADP-TSBP-EC scheme is determined to be predominant based on Classification Accuracy, Precision, Recall and Kappa Statistic in contrast to the standard schemes.

Evaluation of Key Spatiotemporal Learners for Print Track Anomaly Classification Using Melt Pool Image Streams

Recent applications of machine learning in metal additive manufacturing (MAM) have shown great potential to resolve critical barriers to MAM's widespread adoption. Recent research on the topic highlights the significance of using melt pool signatures to predict defects on-the-fly. While high-fidelity melt pool image data has the potential to enable accurate predictions, hardly any work exists on the use of state-of-the-art spatiotemporal models to leverage the information embedded in the transient and sequential nature of the additive process. This work introduces and implements some of the major deep spatiotemporal learners that can be adapted to classify melt pool image streams from different materials, systems, and applications. In this regard, two-stream networks with a spatial and temporal stream, a recurrent spatial network, and a factorized 3D convolutional neural network were investigated herein. The generalization abilities of these models to perturbations in the melt pool image data are tested using data perturbation techniques that are grounded in palpable process situations. The implemented architectures exhibit the ability to learn the spatiotemporal features of the melt pool image sequences. However, only the Kinetics400 pre-trained SlowFast network, which belongs to the two-stream category, showed strong generalisation abilities to data perturbations.

An Intelligent Hybrid Ensemble Gene Selection Model for Autism Using DNN

Autism Spectrum Disorder (ASD) is a complicated neurodevelopmental disorder that is often identified in toddlers. The microarray data is used as a diagnostic tool to identify the genetics of the disorder. However, microarray data is large and has a high volume. Consequently, it suffers from the problem of dimensionality. In microarray data, the sample size and variance of the gene expression will lead to overfitting and misclassification. Identifying the autism gene (feature) subset from microarray data is an important and challenging research area. It has to be efficiently addressed to improve gene feature selection and classification. To overcome the challenges, a novel Intelligent Hybrid Ensemble Gene Selection (IHEGS) model is proposed in this paper. The proposed model integrates the intelligence of different feature selection techniques over the data partitions. In this model, the initial gene selection is carried out by data perturbation, and the final autism gene subset is obtained by functional perturbation, which reduces the problem of dimensionality in microarray data. The functional perturbation module employs three meta-heuristic swarm intelligence-based techniques for gene selection. The obtained gene subset is validated by the Deep Neural Network (DNN) model. The proposed model is implemented using python with six National Center for Biotechnology Information (NCBI) gene expression datasets. From the comparative study with other existing state-of-the-art systems, the proposed model provides stable results in terms of feature selection and classification accuracy.

Использование персистентной энтропии для топологического анализа данных

Persistent homology and persistent entropy have recently become useful tools for pattern recognition. In the paper, requirements are found under which the persistent entropy is stable to small perturbations of the input data and is scale invariant. Stable summary functions are described that combine the persistent entropy and the Betti curve.

EI-HCR: An Efficient End-to-End Hybrid Consistency Regularization Algorithm for Semisupervised Remote Sensing Image Segmentation

Recently, remote sensing image (RSI) semantic segmentation technology has advanced greatly, with the fully supervised process achieving particularly strong performance. However, the technology depends heavily on dataset labels, leading to high annotation costs. To alleviate this problem, we propose a novel efficient end-to-end hybrid consistency regularization algorithm (EI-HCR) for the semi-supervised semantic segmentation of RSI, wherein only a few labeled images and a large number of unlabeled images are effectively used. First, we devise data perturbation (DP) consistency regularization (CR), which includes a data mix-up method to combine unlabeled and labeled images. Then, we employ teacher and student networks to conduct model perturbation (MP) CR. Both segmentation results are regarded as pseudo-labels for each other. In the end, the semi-supervised loss is composed of DP and MP consistency loss, and supervises network training along with the fully supervised loss. More importantly, we first combine the characteristics of knowledge distillation to make the student network more lightweight, efficiently reducing the model inference time. Experimental results demonstrate the effectiveness of EI-HCR on the ISPRS Vaihingen and Massachusetts Buildings datasets. With only 5% of the labeled images, EI-HCR can achieve the same accuracy as the fully supervised training with 50% of the labeled images, and the number of student model parameters is only 9.64 M, indicating the method’s great advantages over other algorithms.