Biblio

With the widespread deployment of data-driven services, the demand for data volumes continues to grow. At present, many applications lack reliable human supervision in the process of data collection, which makes the collected data contain low-quality data or even malicious data. This low-quality or malicious data make AI systems potentially face much security challenges. One of the main security threats in the training phase of machine learning is data poisoning attacks, which compromise model integrity by contaminating training data to make the resulting model skewed or unusable. This paper reviews the relevant researches on data poisoning attacks in various task environments: first, the classification of attacks is summarized, then the defense methods of data poisoning attacks are sorted out, and finally, the possible research directions in the prospect.

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

As the effects of climate change are becoming more and more evident, the importance of improved situation awareness is also gaining more attention, both in the context of preventive environmental monitoring and in the context of acute crisis response. One important aspect of situation awareness is the correct and thorough monitoring of air pollutants. The monitoring is threatened by sensor faults, power or network failures, or other hazards leading to missing or incorrect data transmission. For this reason, in this work we propose two complementary approaches for predicting missing sensor data and a combined technique for detecting outliers. The proposed solution can enhance the performance of low-cost sensor systems, closing the gap of missing measurements due to network unavailability, detecting drift and outliers thus paving the way to its use as an alert system for reportable events. The techniques have been deployed and tested also in a low power microcontroller environment, verifying the suitability of such a computing power to perform the inference locally, leading the way to an edge implementation of a virtual sensor digital twin.

Smart cities are a wide range of projects made to facilitate the problems of everyday life and ensure security. Our interest focuses only on the Intelligent Transport System (ITS) that takes care of the transportation issues using the Vehicular Ad-Hoc Network (VANET) paradigm as its base. VANETs are a promising technology for autonomous driving that provides many benefits to the user conveniences to improve road safety and driving comfort. VANET is a promising technology for autonomous driving that provides many benefits to the user's conveniences by improving road safety and driving comfort. The problem with such rapid development is the continuously increasing digital threats. Among all these threats, we will target the Sybil attack since it has been proved to be one of the most dangerous attacks in VANETs. It allows the attacker to generate multiple forged identities to disseminate numerous false messages, disrupt safety-related services, or misuse the systems. In addition, Machine Learning (ML) is showing a significant influence on classification problems, thus we propose a behavior-based classification algorithm that is tested on the provided VeReMi dataset coupled with various machine learning techniques for comparison. The simulation results prove the ability of our proposed mechanism to detect the Sybil attack in VANETs.

The intrusion detection systems are vital for the sustainability of Cooperative Intelligent Transportation Systems (C-ITS) and the detection of sybil attacks are particularly challenging. In this work, we propose a novel approach for the detection of sybil attacks in C-ITS environments. We provide an evaluation of our approach using extensive simulations that rely on real traces, showing our detection approach's effectiveness.

The Internet of Things (IoT) is advancing technology by creating smart surroundings that make it easier for humans to do their work. This technological advancement not only improves human life and expands economic opportunities, but also allows intruders or attackers to discover and exploit numerous methods in order to circumvent the security of IoT networks. Hence, security and privacy are the key concerns to the IoT networks. It is vital to protect computer and IoT networks from many sorts of anomalies and attacks. Traditional intrusion detection systems (IDS) collect and employ large amounts of data with irrelevant and inappropriate attributes to train machine learning models, resulting in long detection times and a high rate of misclassification. This research presents an advance approach for the design of IDS for IoT networks based on the Particle Swarm Optimization Algorithm (PSO) for feature selection and the Extreme Gradient Boosting (XGB) model for PSO fitness function. The classifier utilized in the intrusion detection process is Random Forest (RF). The IoTID20 is being utilized to evaluate the efficacy and robustness of our suggested strategy. The proposed system attains the following level of accuracy on the IoTID20 dataset for different levels of classification: Binary classification 98 %, multiclass classification 83 %. The results indicate that the proposed framework effectively detects cyber threats and improves the security of IoT networks.

The rapid growth of number of devices that are connected to internet of things (IoT) networks, increases the severity of security problems that need to be solved in order to provide safe environment for network data exchange. The discovery of new vulnerabilities is everyday challenge for security experts and many novel methods for detection and prevention of intrusions are being developed for dealing with this issue. To overcome these shortcomings, artificial intelligence (AI) can be used in development of advanced intrusion detection systems (IDS). This allows such system to adapt to emerging threats, react in real-time and adjust its behavior based on previous experiences. On the other hand, the traffic classification task becomes more difficult because of the large amount of data generated by network systems and high processing demands. For this reason, feature selection (FS) process is applied to reduce data complexity by removing less relevant data for the active classification task and therefore improving algorithm's accuracy. In this work, hybrid version of recently proposed sand cat swarm optimizer algorithm is proposed for feature selection with the goal of increasing performance of extreme learning machine classifier. The performance improvements are demonstrated by validating the proposed method on two well-known datasets - UNSW-NB15 and CICIDS-2017, and comparing the results with those reported for other cutting-edge algorithms that are dealing with the same problems and work in a similar configuration.

This paper introduces an application of machine learning algorithms. In fact, support vector machine and decision tree approaches are studied and applied to compare their performances in detecting, classifying, and locating faults in the transmission network. The IEEE 14-bus transmission network is considered in this work. Besides, 13 types of faults are tested. Particularly, the one fault and the multiple fault cases are investigated and tested separately. Fault simulations are performed using the SimPowerSystems toolbox in Matlab. Basing on the accuracy score, a comparison is made between the proposed approaches while testing simple faults, on the one hand, and when complicated faults are integrated, on the other hand. Simulation results prove that the support vector machine technique can achieve an accuracy of 87% compared to the decision tree which had an accuracy of 53% in complicated cases.

Being a part of today’s technical world, we are connected through a vast network. More we are addicted to these modernization techniques we need security. There must be reliability in a network security system so that it is capable of doing perfect monitoring of the whole network of an organization so that any unauthorized users or intruders wouldn’t be able to halt our security breaches. Firewalls are there for securing our internal network from unauthorized outsiders but still some time possibility of attacks is there as according to a survey 60% of attacks were internal to the network. So, the internal system needs the same higher level of security just like external. So, understanding the value of security measures with accuracy, efficiency, and speed we got to focus on implementing and comparing an improved intrusion detection system. A comprehensive literature review has been done and found that some feature selection techniques with standard scaling combined with Machine Learning Techniques can give better results over normal existing ML Techniques. In this survey paper with the help of the Uni-variate Feature selection method, the selection of 14 essential features out of 41 is performed which are used in comparative analysis. We implemented and compared both binary class classification and multi-class classification-based Intrusion Detection Systems (IDS) for two Supervised Machine Learning Techniques Support Vector Machine and Classification and Regression Techniques.

The major aim of the study is to predict the type of crime that is going to happen based on the crime hotspot detected for the given crime data with engineered spatial features. crime dataset is filtered to have the following 2 crime categories: crime against society, crime against person. Crime hotspots are detected by using the Novel Hierarchical density based Spatial Clustering of Application with Noise (HDBSCAN) Algorithm with the number of clusters optimized using silhouette score. The sample data consists of 501 crime incidents. Future types of crime for the given location are predicted by using the Support Vector Machine (SVM) and Convolutional Neural Network (CNN) algorithms (N=5). The accuracy of crime prediction using Support Vector Machine classification algorithm is 94.01% and Convolutional Neural Network algorithm is 79.98% with the significance p-value of 0.033. The Support Vector Machine algorithm is significantly better in accuracy for prediction of type of crime than Convolutional Neural Network (CNN).

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.

Classifying and predicting the accuracy of intrusion detection on cybercrime by comparing machine learning methods such as Innovative Decision Tree (DT) with Support Vector Machine (SVM). By comparing the Decision Tree (N=20) and the Support Vector Machine algorithm (N=20) two classes of machine learning classifiers were used to determine the accuracy. The decision Tree (99.19%) has the highest accuracy than the SVM (98.5615%) and the independent T-test was carried out (=.507) and shows that it is statistically insignificant (p\textgreater0.05) with a confidence value of 95%. by comparing Innovative Decision Tree and Support Vector Machine. The Decision Tree is more productive than the Support Vector Machine for recognizing intruders with substantially checked, according to the significant analysis.

To solve the current problem of scarce information security talents, this paper proposes to design a network information security attack and defense practical training platform based on ThinkPHP framework. It provides help for areas with limited resources and also offers a communication platform for the majority of information security enthusiasts and students. The platform is deployed using ThinkPHP, and in order to meet the personalized needs of the majority of users, support vector machine algorithms are added to the platform to provide a more convenient service for users.

Malware detection and analysis can be a burdensome task for incident responders. As such, research has turned to machine learning to automate malware detection and malware family classification. Existing work extracts and engineers static and dynamic features from the malware sample to train classifiers. Despite promising results, such techniques assume that the analyst has access to the malware executable file. Self-deleting malware invalidates this assumption and requires analysts to find forensic evidence of malware execution for further analysis. In this paper, we present and evaluate an approach to detecting malware that executed on a Windows target and further classify the malware into its associated family to provide semantic insight. Specifically, we engineer features from the Windows prefetch file, a file system forensic artifact that archives process information. Results show that it is possible to detect the malicious artifact with 99% accuracy; furthermore, classifying the malware into a fine-grained family has comparable performance to techniques that require access to the original executable. We also provide a thorough security discussion of the proposed approach against adversarial diversity.

Due to the rise of the internet a business model known as online advertising has seen unprecedented success. However, it has also become a prime method through which criminals can scam people. Often times even legitimate websites contain advertisements that are linked to scam websites since they are not verified by the website’s owners. Scammers have become quite creative with their attacks, using various unorthodox and inconspicuous methods such as I-frames, Favicons, Proxy servers, Domains, etc. Many modern Anti-viruses are paid services and hence not a feasible option for most users in 3rd world countries. Often people don’t possess devices that have enough RAM to even run such software efficiently leaving them without any options. This project aims to create a Browser extension that will be able to distinguish between safe and unsafe websites by utilizing Machine Learning algorithms. This system is lightweight and free thus fulfilling the needs of most people looking for a cheap and reliable security solution and allowing people to surf the internet easily and safely. The system will scan all the intermittent URL clicks as well, not just the main website thus providing an even greater degree of security.

It is well-known that the most existing machine learning (ML)-based safety-critical applications are vulnerable to carefully crafted input instances called adversarial examples (AXs). An adversary can conveniently attack these target systems from digital as well as physical worlds. This paper aims to the generation of robust physical AXs against face recognition systems. We present a novel smoothness loss function and a patch-noise combo attack for realizing powerful physical AXs. The smoothness loss interjects the concept of delayed constraints during the attack generation process, thereby causing better handling of optimization complexity and smoother AXs for the physical domain. The patch-noise combo attack combines patch noise and imperceptibly small noises from different distributions to generate powerful registration-based physical AXs. An extensive experimental analysis found that our smoothness loss results in robust and more transferable digital and physical AXs than the conventional techniques. Notably, our smoothness loss results in a 1.17 and 1.97 times better mean attack success rate (ASR) in physical white-box and black-box attacks, respectively. Our patch-noise combo attack furthers the performance gains and results in 2.39 and 4.74 times higher mean ASR than conventional technique in physical world white-box and black-box attacks, respectively.

The current adversarial attacks against machine learning models can be divided into white-box attacks and black-box attacks. Further the black-box can be subdivided into soft label and hard label black-box, but the latter has the deficiency of only returning the class with the highest prediction probability, which leads to the difficulty in gradient estimation. However, due to its wide application, it is of great research significance and application value to explore hard label blackbox attacks. This paper proposes an Automatic Selection Attacks Framework (ASAF) for hard label black-box models, which can be explained in two aspects based on the existing attack methods. Firstly, ASAF applies model equivalence to select substitute models automatically so as to generate adversarial examples and then completes black-box attacks based on their transferability. Secondly, specified feature selection and parallel attack method are proposed to shorten the attack time and improve the attack success rate. The experimental results show that ASAF can achieve more than 90% success rate of nontargeted attack on the common models of traditional dataset ResNet-101 (CIFAR10) and InceptionV4 (ImageNet). Meanwhile, compared with FGSM and other attack algorithms, the attack time is reduced by at least 89.7% and 87.8% respectively in two traditional datasets. Besides, it can achieve 90% success rate of attack on the online model, BaiduAI digital recognition. In conclusion, ASAF is the first automatic selection attacks framework for hard label blackbox models, in which specified feature selection and parallel attack methods speed up automatic attacks.

Wireless ad hoc networks are characterized by dynamic topology and high node mobility. Network attacks on wireless ad hoc networks can significantly reduce performance metrics, such as the packet delivery ratio from the source to the destination node, overhead, throughput, etc. The article presents an experimental study of an intrusion detection system prototype in mobile ad hoc networks based on machine learning. The experiment is carried out in a MANET segment of 50 nodes, the detection and prevention of DDoS and cooperative blackhole attacks are investigated. The dependencies of features on the type of network traffic and the dependence of performance metrics on the speed of mobile nodes in the network are investigated. The conducted experimental studies show the effectiveness of an intrusion detection system prototype on simulated data.

Machine Learning (ML) models are now commonly used as components in systems. As any other component, ML components can produce erroneous outputs that may penalize system utility. In this context, self-adaptive systems emerge as a natural approach to cope with ML mispredictions, through the execution of adaptation tactics such as model retraining. To synthesize an adaptation strategy, the self-adaptation manager needs to reason about the cost-benefit tradeoffs of the applicable tactics, which is a non-trivial task for tactics such as model retraining, whose benefits are both context- and data-dependent.To address this challenge, this paper proposes a probabilistic modeling framework that supports automated reasoning about the cost/benefit tradeoffs associated with improving ML components of ML-based systems. The key idea of the proposed approach is to decouple the problems of (i) estimating the expected performance improvement after retrain and (ii) estimating the impact of ML improved predictions on overall system utility.We demonstrate the application of the proposed framework by using it to self-adapt a state-of-the-art ML-based fraud-detection system, which we evaluate using a publicly-available, real fraud detection dataset. We show that by predicting system utility stemming from retraining a ML component, the probabilistic model checker can generate adaptation strategies that are significantly closer to the optimal, as compared against baselines such as periodic retraining, or reactive retraining.

In the world of information technology and the Internet, which has become a part of human life today and is constantly expanding, Attention to the users' requirements such as information security, fast processing, dynamic and instant access, and costs savings has become essential. The solution that is proposed for such problems today is a technology that is called cloud computing. Today, cloud computing is considered one of the most essential distributed tools for processing and storing data on the Internet. With the increasing using this tool, the need to schedule tasks to make the best use of resources and respond appropriately to requests has received much attention, and in this regard, many efforts have been made and are being made. To this purpose, various algorithms have been proposed to calculate resource allocation, each of which has tried to solve equitable distribution challenges while using maximum resources. One of these calculation methods is the DRF algorithm. Although it offers a better approach than previous algorithms, it faces challenges, especially with time-consuming resource allocation computing. These challenges make the use of DRF more complex than ever in the low number of requests with high resource capacity as well as the high number of simultaneous requests. This study tried to reduce the computations costs associated with the DRF algorithm for resource allocation by introducing a new approach to using this DRF algorithm to automate calculations by machine learning and artificial intelligence algorithms (Autonomic Dominant Resource Fairness or A-DRF).

Container security has received much research attention recently. Previous work has proposed to apply various machine learning techniques to detect security attacks in containerized applications. On one hand, supervised machine learning schemes require sufficient labelled training data to achieve good attack detection accuracy. On the other hand, unsupervised machine learning methods are more practical by avoiding training data labelling requirements, but they often suffer from high false alarm rates. In this paper, we present SHIL, a self-supervised hybrid learning solution, which combines unsupervised and supervised learning methods to achieve high accuracy without requiring any manual data labelling. We have implemented a prototype of SHIL and conducted experiments over 41 real world security attacks in 28 commonly used server applications. Our experimental results show that SHIL can reduce false alarms by 39-91% compared to existing supervised or unsupervised machine learning schemes while achieving a higher or similar detection rate.

Machine learning (ML) has been applied in prognostics and health management (PHM) to monitor and predict the health of industrial machinery. The use of PHM in production systems creates a cyber-physical, omni-layer system. While ML offers statistical improvements over previous methods, and brings statistical models to bear on new systems and PHM tasks, it is susceptible to performance degradation when the behavior of the systems that ML is receiving its inputs from changes. Natural changes such as physical wear and engineered changes such as maintenance and rebuild procedures are catalysts for performance degradation, and are both inherent to production systems. Drawing from data on the impact of maintenance procedures on ML performance in hydraulic actuators, this paper presents a simulation study that investigates how long it takes for ML performance degradation to create a difference in the throughput of serial production system. In particular, this investigation considers the performance of an ML model learned on data collected before a rebuild procedure is conducted on a hydraulic actuator and an ML model transfer learned on data collected after the rebuild procedure. Transfer learning is able to mitigate performance degradation, but there is still a significant impact on throughput. The conclusion is drawn that ML faults can have drastic, non-linear effects on the throughput of production systems.

Several excellent devices can communicate without the need for human intervention. It is one of the fastest-growing sectors in the history of computing, with an estimated 50 billion devices sold by the end of 2020. On the one hand, IoT developments play a crucial role in upgrading a few simple, intelligent applications that can increase living quality. On the other hand, the security concerns have been noted to the cross-cutting idea of frameworks and the multidisciplinary components connected with their organization. As a result, encryption, validation, access control, network security, and application security initiatives for gadgets and their inherent flaws cannot be implemented. It should upgrade existing security measures to ensure that the ML environment is sufficiently protected. Machine learning (ML) has advanced tremendously in the last few years. Machine insight has evolved from a research center curiosity to a sensible instrument in a few critical applications.

The model called CSAI-4-CPS is proposed to characterize the use of Artificial Intelligence in Cybersecurity applied to the context of CPS - Cyber-Physical Systems. The model aims to establish a methodology being able to self-adapt using shared machine learning models, without incurring the loss of data privacy. The model will be implemented in a generic framework, to assess accuracy across different datasets, taking advantage of the federated learning and machine learning approach. The proposed solution can facilitate the construction of new AI cybersecurity tools and systems for CPS, enabling a better assessment and increasing the level of security/robustness of these systems more efficiently.

The rapid shift towards smart cities, particularly in the era of pandemics, necessitates the employment of e-learning, remote learning systems, and hybrid models. Building adaptive and personalized education becomes a requirement to mitigate the downsides of distant learning while maintaining high levels of achievement. Explainable artificial intelligence (XAI), machine learning (ML), and the internet of behaviour (IoB) are just a few of the technologies that are helping to shape the future of smart education in the age of smart cities through Customization and personalization. This study presents a paradigm for smart education based on the integration of XAI and IoB technologies. The research uses data acquired on students' behaviours to determine whether or not the current education systems respond appropriately to learners' requirements. Despite the existence of sophisticated education systems, they have not yet reached the degree of development that allows them to be tailored to learners' cognitive needs and support them in the absence of face-to-face instruction. The study collected data on 41 learner's behaviours in response to academic activities and assessed whether the running systems were able to capture such behaviours and respond appropriately or not; the study used evaluation methods that demonstrated that there is a change in students' academic progression concerning monitoring using IoT/IoB to enable a relative response to support their progression.