Adversarial Approach Research Articles

Non-technical losses detection employing adversarial domain adaptation

Non-technical losses (NTL) in power grid not only cause significant financial loss but also reduce the quality of supply for the distribution network operators (DNOs). In recent years, the frequency of power theft activities related to NTL have been increased rapidly. The data-based detection approach for NTLs has been studied by researchers from all over the world thoroughly, but the region domain shift issues and the lack of labelled data limited its applicability. In this paper, an adversarial learning-based approach is proposed to alleviate the domain shift problem, named Cross-Regional Adaptive Network (CAN). First of all, a multilayer Convolutional Neural Networks (CNN) feature extractor maps the training and application scenario data into the same feature space. After that the feature extractor and the domain classifier are trained against each other by inversing extractor’s gradients. Finally, the training and application scenario data are indistinguishable for the domain classifier, and the feature extractor can extract domain invariant features. Experiments are conducted on two sets of electricity consumption data captured by utility company to verify the effectiveness of the proposed method.

A novel state-of-health estimation method for fast charging lithium-ion batteries based on an adversarial encoder network

Deep learning based on big data has been widely implemented with cloud and edge computing for lithium-ion battery online health diagnostics, where enhancing the diagnostics' accuracy, robustness, and real-time application are current research issues. Most research has concentrated on state-of-health estimation methods for diagnostics, but the difference in the degradation feature trajectories of batteries between the training and testing domains has been neglected, which further affects the estimation precision of the trained model. To overcome this issue, a strategy based on adversarial learning and feature selection is proposed. First, the adversarial encoder approach creates a lower-dimensional feature vector to the aggregated domain between the source and target batteries. Then, the feature vector is sent into the estimation module, which consists of a one-dimensional convolutional neural network (1DCNN) to accurately estimate the battery lifespan. The results on the 1C and 6C experimental datasets demonstrate that the adversarial feature selection strategy significantly enhances the estimation performance of the deep learning-based estimation method. Combining an adversarial autoencoder with the prediction module yields statistically superior battery lifetime estimation results.

First Price Sealed-Bid Auctions with Bidders’ Heterogeneous Risk Behavior: An Adversarial Risk Analysis Approach

Bidders’ bidding behavior is analyzed in first price sealed-bid (FPSB) auctions using an adversarial risk analysis (ARA) framework. However, using nonstrategic play and level-k thinking solution concepts, modeling is performed by assuming only two bidders. Also, the aleatory and concept uncertainties have not been yet taken into account by using an ARA framework for these auctions. In this paper, we apply an ARA approach to model bidders’ bidding behavior in a more realistic way for FPSB auctions. We assume n bidders that may have different wealth and heterogeneous risk behaviors. We use nonstrategic play and level-k thinking solution concepts, and we take into account aleatory uncertainty in addition to epistemic uncertainty. Finally, concept uncertainty is taken into account to find ARA solutions for these auctions. We also provide numerical examples to illustrate our methodology.

Bayesian Experimental Design Without Posterior Calculations: An Adversarial Approach

Most computational approaches to Bayesian experimental design require making posterior calculations repeatedly for a large number of potential designs and/or simulated datasets. This can be expensive and prohibit scaling up these methods to models with many parameters, or designs with many unknowns to select. We introduce an efficient alternative approach without posterior calculations, based on optimising the expected trace of the Fisher information, as discussed by Walker (2016). We illustrate drawbacks of this approach, including lack of invariance to reparameterisation and encouraging designs in which one parameter combination is inferred accurately but not any others. We show these can be avoided by using an adversarial approach: the experimenter must select their design while a critic attempts to select the least favourable parameterisation. We present theoretical properties of this approach and show it can be used with gradient based optimisation methods to find designs efficiently in practice.

Network anomaly detection methods in IoT environments via deep learning: A Fair comparison of performance and robustness

The Internet of Things (IoT) is a key enabler in closing the loop in Cyber-Physical Systems, providing “smartness” and thus additional value to each monitored/controlled physical asset. Unfortunately, these devices are more and more targeted by cyberattacks because of their diffusion and of the usually limited hardware and software resources. This calls for designing and evaluating new effective approaches for protecting IoT systems at the network level (Network Intrusion Detection Systems, NIDSs). These in turn are challenged by the heterogeneity of IoT devices and the growing volume of transmitted data.To tackle this challenge, we select a Deep Learning architecture to perform unsupervised early anomaly detection. With a data-driven approach, we explore in-depth multiple design choices and exploit the appealing structural properties of the selected architecture to enhance its performance. The experimental evaluation is performed on two recent and publicly available IoT datasets (IoT-23 and Kitsune). Finally, we adopt an adversarial approach to investigate the robustness of our solution in the presence of Label Flipping poisoning attacks. The experimental results highlight the improved performance of the proposed architecture, in comparison to both well-known baselines and previous proposals.

A data-driven adversarial machine learning for 3D surrogates of unstructured computational fluid dynamic simulations

This paper presents a general workflow to generate and improve the forecast of model surrogates of computational fluid dynamics simulations using deep learning, and most specifically adversarial training. This adversarial approach aims to reduce the divergence of the forecasts from the underlying physical model. Our two-step method integrates a Principal Components Analysis (PCA) based adversarial autoencoder (PC-AAE) with adversarial Long short-term memory (LSTM) networks. Once the reduced-order model (ROM) of the CFD solution is obtained via PCA, an adversarial autoencoder is used on the principal components time series. Subsequentially, a LSTM is adversarially trained on the latent space produced by the PC-AAE to make forecasts. Here we show, that the application of adversarial training improves the rollout of the latent space predictions. Our workflow is applied to three different case studies including two models of urban air pollution in London.

Deep Learning in Asset Pricing

We use deep neural networks to estimate an asset pricing model for individual stock returns that takes advantage of the vast amount of conditioning information, keeps a fully flexible form, and accounts for time variation. The key innovations are to use the fundamental no-arbitrage condition as criterion function to construct the most informative test assets with an adversarial approach and to extract the states of the economy from many macroeconomic time series. Our asset pricing model outperforms out-of-sample all benchmark approaches in terms of Sharpe ratio, explained variation, and pricing errors and identifies the key factors that drive asset prices. This paper was accepted by Agostino Capponi, finance. Supplemental Material: The online appendix and data are available at https://doi.org/10.1287/mnsc.2023.4695 .

An adversarial domain adaptation approach combining dual domain pairing strategy for IoT intrusion detection under few-shot samples

The security of the Internet of Things (IoT) is an integral research topic to support its applicability. With the rapid development of IoT technology, security threats have become increasingly serious. Attackers turn their attention to large IoT devices with many security vulnerabilities, and many infected IoT devices are used to form botnets to launch intrusions. Furthermore, IoT communication networks poses security risks. Therefore, it is necessary to design a model for the IoT intrusion detection. However, the current models are trained using big data samples. Moreover, IoT devices can only intercept a few-shot samples in special scenarios, resulting in lower efficiency and accuracy. In this study, we propose an adversarial domain adaptation (ADDA) approach with dual domain pairing strategy for IoT intrusion detection under few-shot samples. First, we design two data augmentation algorithms to construct a sample transfer model that conforms to the joint dual domain distribution. We then combine the dual domain pairing strategy and ADDA model to perform supervised domain adaptation. In addition, a lightweight network is designed to extract features. Finally, a new loss function and domain discriminator model are constructed to achieve IoT intrusion detection under few-shot samples. Different results demonstrate the effectiveness of our model.

Improving transferable adversarial attack via feature-momentum

Transferable adversarial attackusing adversarial perturbations made on known models to attack unknown modelshas made significant progress in recent years. The feature-level adversarial approach, in particular, is one of the most common solutions and can improve transferability by disrupting intermediate features, regardless of the task-specific loss objectives. Once the intermediate features are disrupted, the subsequent prediction will naturally go wrong. To accomplish this, the existing methods often start an attack by creating a guidance map on features that shows the importance level of each feature element, and then they use an iterative strategy to disrupt the features based on the guidance map. However, the drawback of existing methods is that the guidance map is always fixed in iterations, which can not consistently reflect the importance of feature elements, limiting the performance of the attack consequently. In this paper, we describe a new method called Feature-Momentum Adversarial Attack (FMAA) to enhance transferability. The key idea is that we estimate a guidance map dynamically at each iteration using a momentum-style approach to effectively disturb the features. Extensive experiments demonstrate that our method significantly outperforms other state-of-the-art methods by a large margin on different target models.

Adversarial robustness benchmark for EEG-based brain–computer interfaces

Many machine learning approaches have been successfully applied to electroencephalogram (EEG) based brain–computer interfaces (BCIs). Most existing approaches focused on making EEG-based BCIs more accurate, but few have considered their security. Recent studies have shown that EEG-based BCIs are vulnerable to adversarial attacks. To our knowledge, there is no study on adversarial defenses in EEG-based BCIs. This paper, for the first time, explores multiple classical and state-of-the-art adversarial defense approaches in EEG-based BCIs. Specifically, we demonstrated the performances of nine adversarial defense approaches on three convolutional neural networks and two EEG datasets, and established a comprehensive benchmark to evaluate their effectiveness in BCIs. Based on the evaluation results, we drew some key observations, hoping to inspire future adversarial defense research and attract more attention to the adversarial security of EEG-based BCIs.

What Was the ‘Alt’ in Alt-Right, Alt-Lite, and Alt-Left? On ‘Alt’ as a Political Modifier

What was the ‘alt’ in alt-right, alt-lite, and alt-left? Tracing these controversial terms’ development over the 2010s, this article interrogates the construction, meaning, and utility of the ‘alt’ modifier in US politics. Contextualizing the emergence of ‘alt’ among wider debates about how to spatially conceptualize anti-system politics beyond the left-right spectrum, it argues that ‘alt’ was used to name a tendency that moved beyond traditional ideological questions about what society should look like, and instead signified a preference about how politics should be done: namely, a rejection of the mainstream norms of political conduct, and an embrace of vulgarity, incivility and an extremely adversarial approach to political opposition. It shows how a popular-cultural understanding of ‘alternative’ was used to give this approach a countercultural sheen, and maps out the cleavages between groups that the alt/mainstream binary opened up. Finally, it considers the legacy of the ‘alt’ modifier in contemporary politics.

A bidirectional semantic segmentation method for remote sensing image based on super-resolution and domain adaptation

ABSTRACT Image semantic segmentation methods based on convolutional neural networks rely on supervised learning with labels, and their performance often drops significantly when applied to unlabelled datasets from different sources. The domain adaptation methods can reduce the inconsistency of feature distribution between the unlabelled target domain data used for testing and the labelled source domain data used for training, thus improve the segmentation performance and have more practical applications. However, in the field of remote sensing image processing, if the spatial resolutions of the source domain and the target domain are different and this problem is not to be solved, the performance of the transferred model will be affected. In this paper, we propose a bidirectional semantic segmentation method based on super-resolution and domain adaption (BSSM-SRDA), which is suitable for the transfer learning task of a semantic segmentation model from a low-resolution source domain data to a high-resolution target domain data. BSSM-SRDA mainly consists of three parts: a shared feature extraction network; a super-resolution image translation module, which incorporates a super-resolution approach to reduce spatial resolution differences and visual style differences of the two domains; a domain-adaptive semantic segmentation module, which combines an adversarial domain adaptation approach to reduce differences at the output level. At the same time, we design a new bidirectional self-supervised learning algorithm for BSSM-SRDA that facilitates mutually beneficial learning of the super-resolution image translation module and the domain-adaptive semantic segmentation module. The experiments demonstrate the superiority of the proposed method over other state-of-the-art methods on two remote-sensing image datasets, with mIoU improvements of 2.5% and 3.2%, respectively. Code: https://github.com/mageliang/BSSM-SRDA.git

Towards Backdoor Attacks and Defense in Robust Machine Learning Models

The introduction of robust optimisation has pushed the state-of-the-art in defending against adversarial attacks. Notably, the state-of-the-art projected gradient descent (PGD) -based training method has been shown to be universally and reliably effective in defending against adversarial inputs. This robustness approach uses PGD as a reliable and universal “first-order adversary”. However, the behaviour of such optimisation has not been studied in the light of a fundamentally different class of attacks called backdoors. In this paper, we study how to inject and defend against backdoor attacks for robust models trained using PGD-based robust optimisation. We demonstrate that these models are susceptible to backdoor attacks. Subsequently, we observe that backdoors are reflected in the feature representation of such models. Then, this observation is leveraged to detect such backdoor-infected models via a detection technique called AEGIS. Specifically, given a robust Deep Neural Network (DNN) that is trained using PGD-based first-order adversarial training approach, AEGIS uses feature clustering to effectively detect whether such DNNs are backdoor-infected or clean.In our evaluation of several visible and hidden backdoor triggers on major classification tasks using CIFAR-10, MNIST and FMNIST datasets, AEGIS effectively detects PGD-trained robust DNNs infected with backdoors. AEGIS detects such backdoor-infected models with 91.6% accuracy (11 out of 12 tested models), without any false positives. Furthermore, AEGIS detects the targeted class in the backdoor-infected model with a reasonably low (11.1%) false positive rate. Our investigation reveals that salient features of adversarially robust DNNs could be promising to break the stealthy nature of backdoor attacks.

A Novel Generative Adversarial Network-Based Approach for Automated Brain Tumour Segmentation.

Background: Medical image segmentation is more complicated and demanding than ordinary image segmentation due to the density of medical pictures. A brain tumour is the most common cause of high mortality. Objectives: Extraction of tumorous cells is particularly difficult due to the differences between tumorous and non-tumorous cells. In ordinary convolutional neural networks, local background information is restricted. As a result, previous deep learning algorithms in medical imaging have struggled to detect anomalies in diverse cells. Methods: As a solution to this challenge, a deep convolutional generative adversarial network for tumour segmentation from brain Magnetic resonance Imaging (MRI) images is proposed. A generator and a discriminator are the two networks that make up the proposed model. This network focuses on tumour localisation, noise-related issues, and social class disparities. Results: Dice Score Coefficient (DSC), Peak Signal to Noise Ratio (PSNR), and Structural Index Similarity (SSIM) are all generally 0.894, 62.084 dB, and 0.88912, respectively. The model's accuracy has improved to 97 percent, and its loss has reduced to 0.012. Conclusions: Experiments reveal that the proposed approach may successfully segment tumorous and benign tissues. As a result, a novel brain tumour segmentation approach has been created.

Generative adversarial feature learning for glomerulopathy histological classification

Early diagnosis of glomerulopathy is significant for improving disease control and prognosis. Glomerulopathy is diagnosed by pathological images obtained from a renal biopsy, which has many complications and contraindications. In this work, we propose to utilize noninvasive renal ultrasound images to predict the pathological type of glomerulopathy. Due to the difficulty of pathological type recognition on renal ultrasound images and small amounts of annotated data, effective feature learning is more important. Therefore, we employ an adversarial learning approach to generate feature embeddings rather than images. AC-GAN, using conditional image generation and auxiliary classifier, is deployed to synthesize class-related image features. A basic residual network is applied to extract real renal features. After adversarial feature learning, an additional classifier is used for further optimizing and updating the feature extraction network. Through adversarial learning, the feature embeddings extracted by feature extraction network become more informative and discriminative. The t-SNE visualization of extracted features demonstrates that learned features by adversarial feature learning are more discriminative with image categories than others, and this adversarial feature learning architecture achieves 85% accuracy and 0.8542 AUC value, which is better than single classification network and image generation augmentation strategy.

Fair classification via domain adaptation: A dual adversarial learning approach.

Modern machine learning (ML) models are becoming increasingly popular and are widely used in decision-making systems. However, studies have shown critical issues of ML discrimination and unfairness, which hinder their adoption on high-stake applications. Recent research on fair classifiers has drawn significant attention to developing effective algorithms to achieve fairness and good classification performance. Despite the great success of these fairness-aware machine learning models, most of the existing models require sensitive attributes to pre-process the data, regularize the model learning or post-process the prediction to have fair predictions. However, sensitive attributes are often incomplete or even unavailable due to privacy, legal or regulation restrictions. Though we lack the sensitive attribute for training a fair model in the target domain, there might exist a similar domain that has sensitive attributes. Thus, it is important to exploit auxiliary information from a similar domain to help improve fair classification in the target domain. Therefore, in this paper, we study a novel problem of exploring domain adaptation for fair classification. We propose a new framework that can learn to adapt the sensitive attributes from a source domain for fair classification in the target domain. Extensive experiments on real-world datasets illustrate the effectiveness of the proposed model for fair classification, even when no sensitive attributes are available in the target domain.

Robust Lane Change Decision Making for Autonomous Vehicles: An Observation Adversarial Reinforcement Learning Approach

Reinforcementlearning holds the promise of allowing autonomous vehicles to learn complex decision making behaviors through interacting with other traffic participants. However, many real-world driving tasks involve unpredictable perception errors or measurement noises which may mislead an autonomous vehicle into making unsafe decisions, even cause catastrophic failures. In light of these risks, to ensure safety under perception uncertainty, autonomous vehicles are required to be able to cope with the worst case observation perturbations. Therefore, this paper proposes a novel observation adversarial reinforcement learning approach for robust lane change decision making of autonomous vehicles. A constrained observation-robust Markov decision process is presented to model lane change decision making behaviors of autonomous vehicles under policy constraints and observation uncertainties. Meanwhile, a black-box attack technique based on Bayesian optimization is implemented to approximate the optimal adversarial observation perturbations efficiently. Furthermore, a constrained observation-robust actor-critic algorithm is advanced to optimize autonomous driving lane change policies while keeping the variations of the policies attacked by the optimal adversarial observation perturbations within bounds. Finally, the robust lane change decision making approach is evaluated in three stochastic mixed traffic flows based on different densities. The results demonstrate that the proposed method can not only enhance the performance of an autonomous vehicle but also improve the robustness of lane change policies against adversarial observation perturbations.

Generative Adversarial Networks-Based Novel Approach for Fraud Detection for the European Cardholders 2013 Dataset

Generative Adversarial Networks-Based Novel Approach for Fraud Detection for the European Cardholders 2013 Dataset

A generative adversarial network-based data augmentation approach with transient vibration data

In order to perform fault classification using machine learning algorithms, sufficient and balanced data is required. The paper presents a novel approach for data augmentation using a Generative Adversarial Network (GAN) with transient, time-dependent vibration data. The proposed approach aims to generate synthetic data in form of spectrograms that closely resemble the characteristics of the real data sets. Synthetic data generation can be used to improve the training performance of neural networks for vibration analysis if not enough or unbalanced real data is available. The authors demonstrate the effectiveness of this approach by training a vibration-based fault detection model using synthetic data and comparing its performance to a model trained on real data only as well as with at-distributed stochastic neighbour embedding (t-SNE). Real data is acquired on a bearing test rig and measurements are carried out on bearings with four different system states. The results show that the model trained with the synthetic data set outperforms the model trained with real data only, indicating that the synthetic data generated by the proposed approach can improve the training performance and accuracy of the machine learning model. Overall, the paper highlights the potential of GAN-based data augmentation approach via spectrograms for vibration analysis and offers insights into its practical application to bearings.

Learning Adaptive Probabilistic Models for Uncertainty-Aware Air Pollution Prediction

The air pollution problem has been a serious issue for public health and city development in recent years, which rises an urgent demand for accurate air pollution prediction models. Traditional time series prediction scheme has been challenged in such case because the air pollution signals can be highly dynamic and uncertain. Distinct latent physical processes and complex environment changes over time make it hard for one fixed model to give consistently good performance. Also probabilistic prediction and uncertainty-aware estimation are critical when the model is used for public decisions in practice. However, few previous works can simultaneously provide adaptive and probabilistic air pollution prediction abilities, which are important practical issues for the real-world implementation of data-driven air pollution prediction models. In this paper we propose an adversarial meta learning approach to address these concerns. After reformulating the target signals as a collection of independent but similar mini-tasks along time, we adversarially learn a meta model that includes an implicit generator to adaptively give more informative probabilistic predictions across tasks. We further provide a bayesian meta learning interpretation that recasts the proposed model as an approximate minimizer for the Wasserstein distance between a generative latent model and the true data distribution. Our model provides an algorithm that explicitly models how the prediction distribution is conditioned on underlying data patterns and simultaneously gives adaptive uncertainty estimation. Experiments on both synthetic and real-world air pollution datasets show that the proposed model can simultaneously provide better probabilistic and adaptive prediction while keeping stratifying point prediction error.