Biblio

In recent years, differential privacy has gradually become a standard definition in the field of data privacy protection. Differential privacy does not need to make assumptions about the prior knowledge of privacy adversaries, so it has a more stringent effect than existing privacy protection models and definitions. This good feature has been used by researchers to solve the in-depth learning problem restricted by the problem of privacy and security, making an important breakthrough, and promoting its further large-scale application. Combining differential privacy with BEGAN, we propose the DP-BEGAN framework. The differential privacy is realized by adding carefully designed noise to the gradient of Gan model training, so as to ensure that Gan can generate unlimited synthetic data that conforms to the statistical characteristics of source data and does not disclose privacy. At the same time, it is compared with the existing methods on public datasets. The results show that under a certain privacy budget, this method can generate higher quality privacy protection data more efficiently, which can be used in a variety of data analysis tasks. The privacy loss is independent of the amount of synthetic data, so it can be applied to large datasets.

The smooth operation of metering equipment is inseparable from the monitoring and analysis of equipment alarm events by automated metering systems. With the generation of big data in power metering and the increasing demand for information security of metering systems in the power industry, how to use big data and protect data security at the same time has become a hot research field. In this paper, we propose a hybrid expert model based on federated learning to deal with the problem of alarm information analysis and identification. The hybrid expert system can divide the metering warning problem into multiple sub-problems for processing, which greatly improves the recognition and prediction accuracy. The experimental results show that our model has high accuracy in judging and identifying equipment faults.

Intrusion detection for Controller Area Network (CAN) protocol requires modern methods in order to compete with other electrical architectures. Fingerprint Intrusion Detection Systems (IDS) provide a promising new approach to solve this problem. By characterizing network traffic from known ECUs, hazardous messages can be discriminated. In this article, a modified version of Fingerprint IDS is employed utilizing both step response and spectral characterization of network traffic via neural network training. With the addition of feature set reduction and hyperparameter tuning, this method accomplishes a 99.4% detection rate of trusted ECU traffic.

Nowadays, IoT networks and devices exist in our everyday life, capturing and carrying unlimited data. However, increasing penetration of connected systems and devices implies rising threats for cybersecurity with IoT systems suffering from network attacks. Artificial Intelligence (AI) and Machine Learning take advantage of huge volumes of IoT network logs to enhance their cybersecurity in IoT. However, these data are often desired to remain private. Federated Learning (FL) provides a potential solution which enables collaborative training of attack detection model among a set of federated nodes, while preserving privacy as data remain local and are never disclosed or processed on central servers. While FL is resilient and resolves, up to a point, data governance and ownership issues, it does not guarantee security and privacy by design. Adversaries could interfere with the communication process, expose network vulnerabilities, and manipulate the training process, thus affecting the performance of the trained model. In this paper, we present a federated learning model which can successfully detect network attacks in IoT systems. Moreover, we evaluate its performance under various settings of differential privacy as a privacy preserving technique and configurations of the participating nodes. We prove that the proposed model protects the privacy without actually compromising performance. Our model realizes a limited performance impact of only ∼ 7% less testing accuracy compared to the baseline while simultaneously guaranteeing security and applicability.

Facial is the most dynamic part of the human body that conveys information about emotions. The level of diversity in facial geometry and facial look makes it possible to detect various human expressions. To be able to differentiate among numerous facial expressions of emotion, it is crucial to identify the classes of facial expressions. The methodology used in this article is based on convolutional neural networks (CNN). In this paper Deep Learning CNN is used to examine Alex net architectures. Improvements were achieved by applying the transfer learning approach and modifying the fully connected layer with the Support Vector Machine(SVM) classifier. The system succeeded by achieving satisfactory results on icv-the MEFED dataset. Improved models achieved around 64.29 %of recognition rates for the classification of the selected expressions. The results obtained are acceptable and comparable to the relevant systems in the literature provide ideas a background for further improvements.

Ethical bias in machine learning models has become a matter of concern in the software engineering community. Most of the prior software engineering works concentrated on finding ethical bias in models rather than fixing it. After finding bias, the next step is mitigation. Prior researchers mainly tried to use supervised approaches to achieve fairness. However, in the real world, getting data with trustworthy ground truth is challenging and also ground truth can contain human bias. Semi-supervised learning is a technique where, incrementally, labeled data is used to generate pseudo-labels for the rest of data (and then all that data is used for model training). In this work, we apply four popular semi-supervised techniques as pseudo-labelers to create fair classification models. Our framework, Fair-SSL, takes a very small amount (10%) of labeled data as input and generates pseudo-labels for the unlabeled data. We then synthetically generate new data points to balance the training data based on class and protected attribute as proposed by Chakraborty et al. in FSE 2021. Finally, classification model is trained on the balanced pseudo-labeled data and validated on test data. After experimenting on ten datasets and three learners, we find that Fair-SSL achieves similar performance as three state-of-the-art bias mitigation algorithms. That said, the clear advantage of Fair-SSL is that it requires only 10% of the labeled training data. To the best of our knowledge, this is the first SE work where semi-supervised techniques are used to fight against ethical bias in SE ML models. To facilitate open science and replication, all our source code and datasets are publicly available at https://github.com/joymallyac/FairSSL. CCS CONCEPTS • Software and its engineering → Software creation and management; • Computing methodologies → Machine learning. ACM Reference Format: Joymallya Chakraborty, Suvodeep Majumder, and Huy Tu. 2022. Fair-SSL: Building fair ML Software with less data. In International Workshop on Equitable Data and Technology (FairWare ‘22), May 9, 2022, Pittsburgh, PA, USA. ACM, New York, NY, USA, 8 pages. https://doi.org/10.1145/3524491.3527305

Sybil attack is one of the most dangerous internal attacks in Vehicular Ad Hoc Network (VANET). It affects the function of the VANET network by maliciously claiming or stealing multiple identity propagation error messages. In order to prevent VANET from Sybil attacks, many solutions have been proposed. However, the existing solutions are specific to the physical or application layer's single-level data and lack research on cross-layer information fusion detection. Moreover, these schemes involve a large number of sensitive data access and transmission, do not consider users' privacy, and can also bring a severe communication burden, which will make these schemes unable to be actually implemented. In this context, this paper introduces FedMix, the first federated Sybil attack detection system that considers cross-layer information fusion and provides privacy protection. The system can integrate VANET physical layer data and application layer data for joint analyses simultaneously. The data resides locally in the vehicle for local training. Then, the central agency only aggregates the generated model and finally distributes it to the vehicles for attack detection. This process does not involve transmitting and accessing any vehicle's original data. Meanwhile, we also designed a new model aggregation algorithm called SFedAvg to solve the problems of unbalanced vehicle data quality and low aggregation efficiency. Experiments show that FedMix can provide an intelligent model with equivalent performance under the premise of privacy protection and significantly reduce communication overhead, compared with the traditional centralized training attack detection model. In addition, the SFedAvg algorithm and cross-layer information fusion bring better aggregation efficiency and detection performance, respectively.

In recent years, the security of AI systems has drawn increasing research attention, especially in the medical imaging realm. To develop a secure medical image analysis (MIA) system, it is a must to study possible backdoor attacks (BAs), which can embed hidden malicious behaviors into the system. However, designing a unified BA method that can be applied to various MIA systems is challenging due to the diversity of imaging modalities (e.g., X-Ray, CT, and MRI) and analysis tasks (e.g., classification, detection, and segmentation). Most existing BA methods are designed to attack natural image classification models, which apply spatial triggers to training images and inevitably corrupt the semantics of poisoned pixels, leading to the failures of attacking dense prediction models. To address this issue, we propose a novel Frequency-Injection based Backdoor Attack method (FIBA) that is capable of delivering attacks in various MIA tasks. Specifically, FIBA leverages a trigger function in the frequency domain that can inject the low-frequency information of a trigger image into the poisoned image by linearly combining the spectral amplitude of both images. Since it preserves the semantics of the poisoned image pixels, FIBA can perform attacks on both classification and dense prediction models. Experiments on three benchmarks in MIA (i.e., ISIC-2019 [4] for skin lesion classification, KiTS-19 [17] for kidney tumor segmentation, and EAD-2019 [1] for endoscopic artifact detection), validate the effectiveness of FIBA and its superiority over stateof-the-art methods in attacking MIA models and bypassing backdoor defense. Source code will be available at code.

This paper aims to contribute towards creating ambient abnormal behavior detection for smart security system from real-time human pose estimation using fuzzy-based systems. Human poses from keypoint detected by pose estimation model are transformed to as angle positions of the axis between human bodies joints comparing to reference point in the axis x to deal with problem of the position change occurred when an individual move in the image. Also, the article attempts to resolve the problem of the ambiguity interpreting the poses with triangular fuzzy logic-based system that determines the detected individual behavior and compares to the poses previously learnt, trained, and recorded by the system. The experiment reveals that the accuracy of the system ranges between 90.75% (maximum) and 84% (minimum). This means that if the accuracy of the system at 85%. The system can be applied to guide future research for designing automatic visual human behavior detection systems.

Recently exposed vulnerabilities reveal the necessity to improve the security of branch predictors. Branch predictors record history about the execution of different processes, and such information from different processes are stored in the same structure and thus accessible to each other. This leaves the attackers with the opportunities for malicious training and malicious perception. Physical or logical isolation mechanisms such as using dedicated tables and flushing during context-switch can provide security but incur non-trivial costs in space and/or execution time. Randomization mechanisms incurs the performance cost in a different way: those with higher securities add latency to the critical path of the pipeline, while the simpler alternatives leave vulnerabilities to more sophisticated attacks.This paper proposes HyBP, a practical hybrid protection and effective mechanism for building secure branch predictors. The design applies the physical isolation and randomization in the right component to achieve the best of both worlds. We propose to protect the smaller tables with physically isolation based on (thread, privilege) combination; and protect the large tables with randomization. Surprisingly, the physical isolation also significantly enhances the security of the last-level tables by naturally filtering out accesses, reducing the information flow to these bigger tables. As a result, key changes can happen less frequently and be performed conveniently at context switches. Moreover, we propose a latency hiding design for a strong cipher by precomputing the "code book" with a validated, cryptographically strong cipher. Overall, our design incurs a performance penalty of 0.5% compared to 5.1% of physical isolation under the default context switching interval in Linux.

Graph-based Semi-Supervised Learning (GSSL) is a practical solution to learn from a limited amount of labelled data together with a vast amount of unlabelled data. However, due to their reliance on the known labels to infer the unknown labels, these algorithms are sensitive to data quality. It is therefore essential to study the potential threats related to the labelled data, more specifically, label poisoning. In this paper, we propose a novel data poisoning method which efficiently approximates the result of label inference to identify the inputs which, if poisoned, would produce the highest number of incorrectly inferred labels. We extensively evaluate our approach on three classification problems under 24 different experimental settings each. Compared to the state of the art, our influence-driven attack produces an average increase of error rate 50% higher, while being faster by multiple orders of magnitude. Moreover, our method can inform engineers of inputs that deserve investigation (relabelling them) before training the learning model. We show that relabelling one-third of the poisoned inputs (selected based on their influence) reduces the poisoning effect by 50%. ACM Reference Format: Adriano Franci, Maxime Cordy, Martin Gubri, Mike Papadakis, and Yves Le Traon. 2022. Influence-Driven Data Poisoning in Graph-Based Semi-Supervised Classifiers. In 1st Conference on AI Engineering - Software Engineering for AI (CAIN’22), May 16–24, 2022, Pittsburgh, PA, USA. ACM, New York, NY, USA, 11 pages. https://doi.org/10.1145/3522664.3528606

Transformer is the key equipment of power system, and its stable operation is very important to the security of power system In practical application, with the progress of technology, the performance of transformer becomes more and more important, but faults also occur from time to time in practical application, and the traditional manual fault diagnosis needs to consume a lot of time and energy. At present, the rapid development of artificial intelligence technology provides a new research direction for timely and accurate detection and treatment of transformer faults. In this paper, a method of transformer fault diagnosis using artificial neural network is proposed. The neural network algorithm is used for off-line learning and training of the operation state data of normal and fault states. By adjusting the relationship between neuron nodes, the mapping relationship between fault characteristics and fault location is established by using network layer learning, Finally, the reasoning process from fault feature to fault location is realized to realize intelligent fault diagnosis.

The compressed sensing (CS) method can reconstruct images with a small amount of under-sampling data, which is an effective method for fast magnetic resonance imaging (MRI). As the traditional optimization-based models for MRI suffered from non-adaptive sampling and shallow” representation ability, they were unable to characterize the rich patterns in MRI data. In this paper, we propose a CS MRI method based on iterative shrinkage threshold algorithm (ISTA) and adaptive sparse sampling, called DSLS-ISTA-Net. Corresponding to the sampling and reconstruction of the CS method, the network framework includes two folders: the sampling sub-network and the improved ISTA reconstruction sub-network which are coordinated with each other through end-to-end training in an unsupervised way. The sampling sub-network and ISTA reconstruction sub-network are responsible for the implementation of adaptive sparse sampling and deep sparse representation respectively. In the testing phase, we investigate different modules and parameters in the network structure, and perform extensive experiments on MR images at different sampling rates to obtain the optimal network. Due to the combination of the advantages of the model-based method and the deep learning-based method in this method, and taking both adaptive sampling and deep sparse representation into account, the proposed networks significantly improve the reconstruction performance compared to the art-of-state CS-MRI approaches.

Unsupervised cross-domain NER task aims to solve the issues when data in a new domain are fully-unlabeled. It leverages labeled data from source domain to predict entities in unlabeled target domain. Since training models on large domain corpus is time-consuming, in this paper, we consider an alternative way by introducing syntactic dependency structure. Such information is more accessible and can be shared between sentences from different domains. We propose a novel framework with dependency-aware GNN (DGNN) to learn these common structures from source domain and adapt them to target domain, alleviating the data scarcity issue and bridging the domain gap. Experimental results show that our method outperforms state-of-the-art methods.

Detection of malware and security attacks is a complex process that can vary in its details and analysis activities. As part of the detection process, malware scanners try to categorize a malware once it is detected under one of the known malware categories (e.g. worms, spywares, viruses, etc.). However, many studies and researches indicate problems with scanners categorizing or identifying a particular malware under more than one malware category. This paper, and several others, show that machine learning can be used for malware detection especially with ensemble base prediction methods. In this paper, we evaluated several custom-built ensemble models. We focused on multi-label malware classification as individual or classical classifiers showed low accuracy in such territory.This paper showed that recent machine models such as ensemble and deep learning can be used for malware detection with better performance in comparison with classical models. This is very critical in such a dynamic and yet important detection systems where challenges such as the detection of unknown or zero-day malware will continue to exist and evolve.

The development of industrial automation tools and the integration of industrial and corporate networks in order to improve the quality of production management have led to an increase in the risks of successful cyberattacks and, as a result, to the necessity to solve the problems of practical information security of industrial control systems (ICS). Detection of cyberattacks of both known and unknown types is could be implemented as anomaly detection in dynamic information processes recorded during the operation of ICS. Anomaly detection methods do not require preliminary analysis and labeling of the training sample. In the context of detecting attacks on ICS, cluster analysis is used as one of the methods that implement anomaly detection. The application of hierarchical cluster analysis for clustering data of ICS information processes exposed to various cyberattacks is studied, the problem of choosing the level of the cluster hierarchy corresponding to the minimum set of clusters aggregating separately normal and abnormal data is solved. It is shown that the Ward method of hierarchical cluster division produces the best division into clusters. The next stage of the study involves solving the problem of classifying the formed minimum set of clusters, that is, determining which cluster is normal and which cluster is abnormal.

The security of embedded systems is particularly crucial given the prevalence of embedded devices in daily life, business, and national defense. Firmware for embedded systems poses a serious threat to the safety of society, business, and the nation because of its robust concealment, difficulty in detection, and extended maintenance cycle. This technology is now an essential part of the contemporary experience, be it in the smart office, smart restaurant, smart home, or even the smart traffic system. Despite the fact that these systems are often fairly effective, the rapid expansion of embedded systems in smart cities have led to inconsistencies and misalignments between secured and unsecured systems, necessitating the development of secure, hacker-proof embedded systems. To solve this issue, we created a sizable, original, and objective dataset that is based on the latest Linux vulnerabilities for identifying the embedded system vulnerabilities and we modified a cutting-edge machine learning model for the Linux Kernel. The paper provides an updated EVDD and analysis of an extensive dataset for embedded system based vulnerability detection and also an updated state of the art deep learning model for embedded system vulnerability detection. We kept our dataset available for all researchers for future experiments and implementation.

We introduce AdaMix, an adaptive differentially private algorithm for training deep neural network classifiers using both private and public image data. While pre-training language models on large public datasets has enabled strong differential privacy (DP) guarantees with minor loss of accuracy, a similar practice yields punishing trade-offs in vision tasks. A few-shot or even zero-shot learning baseline that ignores private data can outperform fine-tuning on a large private dataset. AdaMix incorporates few-shot training, or cross-modal zero-shot learning, on public data prior to private fine-tuning, to improve the trade-off. AdaMix reduces the error increase from the non-private upper bound from the 167–311% of the baseline, on average across 6 datasets, to 68-92% depending on the desired privacy level selected by the user. AdaMix tackles the trade-off arising in visual classification, whereby the most privacy sensitive data, corresponding to isolated points in representation space, are also critical for high classification accuracy. In addition, AdaMix comes with strong theoretical privacy guarantees and convergence analysis.

State of the art Artificial Intelligence Assurance (AIA) methods validate AI systems based on predefined goals and standards, are applied within a given domain, and are designed for a specific AI algorithm. Existing works do not provide information on assuring subjective AI goals such as fairness and trustworthiness. Other assurance goals are frequently required in an intelligent deployment, including explainability, safety, and security. Accordingly, issues such as value loading, generalization, context, and scalability arise; however, achieving multiple assurance goals without major trade-offs is generally deemed an unattainable task. In this manuscript, we present two AIA pipelines that are model-agnostic, independent of the domain (such as: healthcare, energy, banking), and provide scores for AIA goals including explainability, safety, and security. The two pipelines: Adversarial Logging Scoring Pipeline (ALSP) and Requirements Feedback Scoring Pipeline (RFSP) are scalable and tested with multiple use cases, such as a water distribution network and a telecommunications network, to illustrate their benefits. ALSP optimizes models using a game theory approach and it also logs and scores the actions of an AI model to detect adversarial inputs, and assures the datasets used for training. RFSP identifies the best hyper-parameters using a Bayesian approach and provides assurance scores for subjective goals such as ethical AI using user inputs and statistical assurance measures. Each pipeline has three algorithms that enforce the final assurance scores and other outcomes. Unlike ALSP (which is a parallel process), RFSP is user-driven and its actions are sequential. Data are collected for experimentation; the results of both pipelines are presented and contrasted.

This paper is a conceptual paper that explores how the sensemaking process by intelligence analysts completed within a cognitive immersive environment might be impacted by the inclusion of a progressive dialog system. The tools enabled in the sensemaking room (a specific instance within the cognitive immersive environment) were informed by tools from the intelligence analysis domain. We explore how a progressive dialog system would impact the use of tools such as the collaborative brainstorming exercise [1]. These structured analytic techniques are well established in intelligence analysis training literature, and act as ways to access the intended users' cognitive schema as they use the cognitive immersive room and move through the sensemaking process. A prior user study determined that the sensemaking room encouraged users to be more concise and representative with information while using the digital brainstorming tool. We anticipate that addition of the progressive dialog function will enable a more cohesive link between information foraging and sensemaking behaviors for analysts.

With the development of industrial informatization, information security in the power production industry is becoming more and more important. In the power production industry, as the critical information egress of the industrial control system, the information security of the Networked Control System is particularly important. This paper proposes a construction method for an information security platform of Networked Control System, which is used for research, testing and training of Networked Control System information security.

Higher education management has problems producing 100% of graduates capable of responding to the needs of industry while industry also is struggling to find qualified graduates that responded to their needs in part because of the inefficient way of evaluating problems, as well as because of weaknesses in the evaluation of problem-solving capabilities. The objective of this paper is to propose an appropriate classification model to be used for predicting and evaluating the attributes of the data set of the student in order to meet the selection criteria required by the industries in the academic field. The dataset required for this analysis was obtained from a private firm and the execution was carried out using Chimp Optimization Algorithm (COA) based Deep Belief Neural Network (COA-DBNN) and the obtained results are compared with various classifiers such as Logistic Regression (LR), Decision Tree (DT) and Random Forest (RF). The proposed model outperforms other classifiers in terms of various performance metrics. This critical analysis will help the college management to make a better long-term plan for producing graduates who are skilled, knowledgeable and fulfill the industry needs as well.

With the progress of cryptography computer science, designing cryptographic algorithms using deep learning is a very innovative research direction. Google Brain designed a communication model using generation adversarial network and explored the encrypted communication algorithm based on machine learning. However, the encrypted communication model it designed lacks quantitative evaluation. When some plaintexts and keys are leaked at the same time, the security of communication cannot be guaranteed. This model is optimized to enhance the security by adjusting the optimizer, modifying the activation function, and increasing batch normalization to improve communication speed of optimization. Experiments were performed on 16 bits and 64 bits plaintexts communication. With plaintext and key leak rate of 0.75, the decryption error rate of the decryptor is 0.01 and the attacker can't guess any valid information about the communication.

Many organizations use internal phishing campaigns to gauge awareness and coordinate training efforts based on those findings. Ongoing content design is important for phishing training tools due to the influence recency has on phishing susceptibility. Traditional approaches for content development require significant investment and can be prohibitively costly, especially during the requirements engineering phase of software development and for applications that are constantly evolving. While prior research primarily depends upon already known phishing cues curated by experts, our project, Phish Finders, uses crowdsourcing to explore phishing cues through the unique perspectives and thought processes of everyday users in a realistic yet safe online environment, Zooniverse. This paper contributes qualitative analysis of crowdsourced comments that identifies novel cues, such as formatting and typography, which were identified by the crowd as potential phishing indicators. The paper also shows that crowdsourcing may have the potential to scale as a requirements engineering approach to meet the needs of content labeling for improved training tool development.

A persistent and serious danger to the Internet is a denial of service attack on a large scale (DDoS) attack using machine learning. Because they originate at the low layers, new Infections that use genuine hypertext transfer protocol requests to overload target resources are more untraceable than application layer-based cyberattacks. Using network flow traces to construct an access matrix, this research presents a method for detecting distributed denial of service attack machine learning assaults. Independent component analysis decreases the number of attributes utilized in detection because it is multidimensional. Independent component analysis can be used to translate features into high dimensions and then locate feature subsets. Furthermore, during the training and testing phase of the updated source support vector machine for classification, their performance it is possible to keep track of the detection rate and false alarms. Modified source support vector machine is popular for pattern classification because it produces good results when compared to other approaches, and it outperforms other methods in testing even when given less information about the dataset. To increase classification rate, modified source support Vector machine is used, which is optimized using BAT and the modified Cuckoo Search method. When compared to standard classifiers, the acquired findings indicate better performance.