Biblio

Data poisoning is a type of adversarial attack on training data where an attacker manipulates a fraction of data to degrade the performance of machine learning model. There are several known defensive mechanisms for handling offline attacks, however defensive measures for online learning, where data points arrive sequentially, have not garnered similar interest. In this work, we propose a defense mechanism to minimize the degradation caused by the poisoned training data on a learner's model in an online setup. Our proposed method utilizes an influence function which is a classic technique in robust statistics. Further, we supplement it with the existing data sanitization methods for filtering out some of the poisoned data points. We study the effectiveness of our defense mechanism on multiple datasets and across multiple attack strategies against an online learner.

In this paper, stock selection strategy design based on machine learning and multi-factor analysis is a research hotspot in quantitative investment field. Four machine learning algorithms including support vector machine, gradient lifting regression, random forest and linear regression are used to predict the rise and fall of stocks by taking stock fundamentals as input variables. The portfolio strategy is constructed on this basis. Finally, the stock selection strategy is further optimized. The empirical results show that the multifactor quantitative stock selection strategy has a good stock selection effect, and yield performance under the support vector machine algorithm is the best. With the increase of the number of factors, there is an inverse relationship between the fitting degree and the yield under various algorithms.

Facial expression intensity, which quantifies the degree of facial expression, has been proposed. It is calculated based on how much facial feature values change compared to an expressionless face. The estimation has two aspects. One is to classify facial expressions, and the other is to estimate their intensity. However, it is difficult to do them at the same time. There- fore, in this work, the estimation of intensity and the classification of expression are separated. We suggest an explicit method and an implicit method. In the explicit one, a classifier determines which types of expression the inputs are, and each regressor determines its intensity. On the other hand, in the implicit one, we give zero values or non-zero values to regressors for each type of facial expression as ground truth, depending on whether or not an input image is the correct facial expression. We evaluated the two methods and, as a result, found that they are effective for facial expression recognition.

A good ecological environment is crucial to attracting talents, cultivating talents, retaining talents and making talents fully effective. This study provides a solution to the current mainstream problem of how to deal with excellent employee turnover in advance, so as to promote the sustainable and harmonious human resources ecological environment of enterprises with a shortage of talents.This study obtains open data sets and conducts data preprocessing, model construction and model optimization, and describes a set of enterprise employee turnover prediction models based on RapidMiner workflow. The data preprocessing is completed with the help of the data statistical analysis software IBM SPSS Statistic and RapidMiner.Statistical charts, scatter plots and boxplots for analysis are generated to realize data visualization analysis. Machine learning, model application, performance vector, and cross-validation through RapidMiner's multiple operators and workflows. Model design algorithms include support vector machines, naive Bayes, decision trees, and neural networks. Comparing the performance parameters of the algorithm model from the four aspects of accuracy, precision, recall and F1-score. It is concluded that the performance of the decision tree algorithm model is the highest. The performance evaluation results confirm the effectiveness of this model in sustainable exploring of enterprise employee turnover prediction in human resource management.

Most of the recent high-profile attacks targeting cyber-physical systems (CPS) started with lengthy reconnaissance periods that enabled attackers to gain in-depth understanding of the victim’s environment. To simulate these stealthy attacks, several covert channel tools have been published and proven effective in their ability to blend into existing CPS communication streams and have the capability for data exfiltration and command injection.In this paper, we report a novel machine learning feature engineering and data processing pipeline for the detection of covert channel attacks on CPS systems with real-time detection throughput. The system also operates at the network layer without requiring physical system domain-specific state modeling, such as voltage levels in a power generation system. We not only demonstrate the effectiveness of using TCP payload entropy as engineered features and the technique of grouping information into network flows, but also pitch the proposed detector against scenarios employing advanced evasion tactics, and still achieve above 99% detection performance.

Cybersecurity insurance is one of the important means of cybersecurity risk management and the development of cyber insurance is inseparable from the support of cyber risk assessment technology. Cyber risk assessment can not only help governments and organizations to better protect themselves from related risks, but also serve as a basis for cybersecurity insurance underwriting, pricing, and formulating policy content. Aiming at the problem that cybersecurity insurance companies cannot conduct cybersecurity risk assessments on policyholders before the policy is signed without the authorization of the policyholder or in legal, combining with the need that cybersecurity insurance companies want to obtain network security vulnerability risk profiles of policyholders conveniently, quickly and at low cost before the policy signing, this study proposed a non-intrusive network security vulnerability risk assessment method based on ensemble machine learning. Our model uses only open source intelligence and publicly available network information data to rate cyber vulnerability risk of an organization, achieving an accuracy of 70.6% compared to a rating based on comprehensive information by cybersecurity experts.

Artificial intelligence creation comes into fashion and has brought unprecedented challenges to intellectual property law. In order to study the viewpoints of AI creation copyright ownership from professionals in different institutions, taking the papers of AI creation on CNKI from 2016 to 2021, we applied orthogonal design and analysis of variance method to construct the dataset. A kernel-SVM classifier with different kernel methods in addition to some shallow machine learning classifiers are selected in analyzing and predicting the copyright ownership of AI creation. Support vector machine (svm) is widely used in statistics and the performance of SVM method is closely related to the choice of the kernel function. SVM with RBF kernel surpasses the other seven kernel-SVM classifiers and five shallow classifier, although the accuracy provided by all of them was not satisfactory. Various performance metrics such as accuracy, F1-score are used to evaluate the performance of KSVM and other classifiers. The purpose of this study is to explore the overall viewpoints of AI creation copyright ownership, investigate the influence of different features on the final copyright ownership and predict the most likely viewpoint in the future. And it will encourage investors, researchers and promote intellectual property protection in China.

Nowadays, network information security is of great concern, and the measurement of the trustworthiness of terminal devices is of great significance to the security of the entire network. The measurement method of terminal device security trust still has the problems of high complexity, lack of universality. In this paper, the device fingerprint library of device access network terminal devices is first established through the device fingerprint mixed collection method; Secondly, the software and hardware features of the device fingerprint are used to increase the uniqueness of the device identification, and the multi- dimensional standard metric is used to measure the trustworthiness of the terminal device; Finally, Block chain technology is used to store the fingerprint and standard model of network access terminal equipment on the chain. To improve the security level of network access devices, a device access method considering the trust of terminal devices from multiple perspectives is implemented.

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.

Current intrusion detection techniques cannot keep up with the increasing amount and complexity of cyber attacks. In fact, most of the traffic is encrypted and does not allow to apply deep packet inspection approaches. In recent years, Machine Learning techniques have been proposed for post-mortem detection of network attacks, and many datasets have been shared by research groups and organizations for training and validation. Differently from the vast related literature, in this paper we propose an early classification approach conducted on CSE-CIC-IDS2018 dataset, which contains both benign and malicious traffic, for the detection of malicious attacks before they could damage an organization. To this aim, we investigated a different set of features, and the sensitivity of performance of five classification algorithms to the number of observed packets. Results show that ML approaches relying on ten packets provide satisfactory results.

Black-box adversarial attack has aroused much research attention for its difficulty on nearly no available information of the attacked model and the additional constraint on the query budget. A common way to improve attack efficiency is to transfer the gradient information of a white-box substitute model trained on an extra dataset. In this paper, we deal with a more practical setting where a pre-trained white-box model with network parameters is provided without extra training data. To solve the model mismatch problem between the white-box and black-box models, we propose a novel algorithm EigenBA by systematically integrating gradient-based white-box method and zeroth-order optimization in black-box methods. We theoretically show the optimal directions of perturbations for each step are closely related to the right singular vectors of the Jacobian matrix of the pretrained white-box model. Extensive experiments on ImageNet, CIFAR-10 and WebVision show that EigenBA can consistently and significantly outperform state-of-the-art baselines in terms of success rate and attack efficiency.

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.

Global cybersecurity threats have grown as a result of the evolving digital transformation. Cybercriminals have more opportunities as a result of digitization. Initially, cyberthreats take the form of phishing in order to gain confidential user credentials.As cyber-attacks get more sophisticated and sophisticated, the cybersecurity industry is faced with the problem of utilising cutting-edge technology and techniques to combat the ever-present hostile threats. Hackers use phishing to persuade customers to grant them access to a company’s digital assets and networks. As technology progressed, phishing attempts became more sophisticated, necessitating the development of tools to detect phishing.Machine learning is unsupervised one of the most powerful weapons in the fight against terrorist threats. The features used for phishing detection, as well as the approaches employed with machine learning, are discussed in this study.In this light, the study’s major goal is to propose a unique, robust ensemble machine learning model architecture that gives the highest prediction accuracy with the lowest error rate, while also recommending a few alternative robust machine learning models.Finally, the Random forest algorithm attained a maximum accuracy of 96.454 percent. But by implementing a hybrid model including the 3 classifiers- Decision Trees,Random forest, Gradient boosting classifiers, the accuracy increases to 98.4 percent.

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.

This paper presents the machine learning algorithm to detect whether an executable binary is benign or ransomware. The ransomware cybercriminals have targeted our infrastructure, businesses, and everywhere which has directly affected our national security and daily life. Tackling the ransomware threats more effectively is a big challenge. We applied a machine-learning model to classify and identify the security level for a given suspected malware for ransomware detection and prevention. We use the feature selection data preprocessing to improve the prediction accuracy of the model.

For a long time, SQL injection has been considered one of the most serious security threats. NoSQL databases are becoming increasingly popular as big data and cloud computing technologies progress. NoSQL injection attacks are designed to take advantage of applications that employ NoSQL databases. NoSQL injections can be particularly harmful because they allow unrestricted code execution. In this paper we use supervised learning and natural language processing to construct a model to detect NoSQL injections. Our model is designed to work with MongoDB, CouchDB, CassandraDB, and Couchbase queries. Our model has achieved an F1 score of 0.95 as established by 10-fold cross validation.

During pandemic COVID-19 outbreaks, number of cyber-attacks including phishing activities have increased tremendously. Nowadays many technical solutions on phishing detection were developed, however these approaches were either unsuccessful or unable to identify phishing pages and detect malicious codes efficiently. One of the downside is due to poor detection accuracy and low adaptability to new phishing connections. Another reason behind the unsuccessful anti-phishing solutions is an arbitrary selected URL-based classification features which may produce false results to the detection. Therefore, in this work, an intelligent phishing detection and prevention model is designed. The proposed model employs a self-destruct detection algorithm in which, machine learning, especially supervised learning algorithm was used. All employed rules in algorithm will focus on URL-based web characteristic, which attackers rely upon to redirect the victims to the simulated sites. A dataset from various sources such as Phish Tank and UCI Machine Learning repository were used and the testing was conducted in a controlled lab environment. As a result, a chrome extension phishing detection were developed based on the proposed model to help in preventing phishing attacks with an appropriate countermeasure and keep users aware of phishing while visiting illegitimate websites. It is believed that this smart phishing detection and prevention model able to prevent fraud and spam websites and lessen the cyber-crime and cyber-crisis that arise from year to year.

Ransomware uses encryption methods to make data inaccessible to legitimate users. To date a wide range of ransomware families have been developed and deployed, causing immense damage to governments, corporations, and private users. As these cyberthreats multiply, researchers have proposed a range of ransom ware detection and classification schemes. Most of these methods use advanced machine learning techniques to process and analyze real-world ransomware binaries and action sequences. Hence this paper presents a survey of this critical space and classifies existing solutions into several categories, i.e., including network-based, host-based, forensic characterization, and authorship attribution. Key facilities and tools for ransomware analysis are also presented along with open challenges.

Machine Learning (ML) and Artificial Intelligence (AI) techniques are widely adopted in the telecommunication industry, especially to automate beyond 5G networks. Federated Learning (FL) recently emerged as a distributed ML approach that enables localized model training to keep data decentralized to ensure data privacy. In this paper, we identify the applicability of FL for securing future networks and its limitations due to the vulnerability to poisoning attacks. First, we investigate the shortcomings of state-of-the-art security algorithms for FL and perform an attack to circumvent FoolsGold algorithm, which is known as one of the most promising defense techniques currently available. The attack is launched with the addition of intelligent noise at the poisonous model updates. Then we propose a more sophisticated defense strategy, a threshold-based clustering mechanism to complement FoolsGold. Moreover, we provide a comprehensive analysis of the impact of the attack scenario and the performance of the defense mechanism.

In recent times, the occurrence of malware attacks are increasing at an unprecedented rate. Particularly, the image-based malware attacks are spreading worldwide and many people get harmful malware-based images through the technique called steganography. In the existing system, only open malware and files from the internet can be identified. However, the image-based malware cannot be identified and detected. As a result, so many phishers make use of this technique and exploit the target. Social media platforms would be totally harmful to the users. To avoid these difficulties, Machine learning can be implemented to find the steganographic malware images (contents). The proposed methodology performs an automatic detection of malware and steganographic content by using Machine Learning. Steganography is used to hide messages from apparently innocuous media (e.g., images), and steganalysis is the approach used for detecting this malware. This research work proposes a machine learning (ML) approach to perform steganalysis. In the existing system, only open malware and files from the internet are identified but in the recent times many people get harmful malware-based images through the technique called steganography. Social media platforms would be totally harmful to the users. To avoid these difficulties, the proposed Machine learning has been developed to appropriately detect the steganographic malware images (contents). Father, the steganalysis method using machine learning has been developed for performing logistic classification. By using this, the users can avoid sharing the malware images in social media platforms like WhatsApp, Facebook without downloading it. It can be also used in all the photo-sharing sites such as google photos.

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.

Binary analysis is pervasively utilized to assess software security and test vulnerabilities without accessing source codes. The analysis validity is heavily influenced by the inferring ability of information related to the code compilation. Among the compilation information, compiler type and optimization level, as the key factors determining how binaries look like, are still difficult to be inferred efficiently with existing tools. In this paper, we conduct a thorough empirical study on the binary's appearance under various compilation settings and propose a lightweight binary analysis tool based on the simplest machine learning method, called DIComP to infer the compiler and optimization level via most relevant features according to the observation. Our comprehensive evaluations demonstrate that DIComP can fully recognize the compiler provenance, and it is effective in inferring the optimization levels with up to 90% accuracy. Also, it is efficient to infer thousands of binaries at a millisecond level with our lightweight machine learning model (1MB).

Community question answering (CQA) websites have become very popular platforms attracting numerous participants to share and acquire knowledge and information in Internet However, with the rapid growth of crowdsourcing systems, many malicious users organize collusive attacks against the CQA platforms for promoting a target (product or service) via posting suggestive questions and deceptive answers. These manipulate deceptive contents, aggregating into multiple collusive questions and answers (Q&As) spam groups, can fully control the sentiment of a target and distort the decision of users, which pollute the CQA environment and make it less credible. In this paper, we propose a Pattern and Burstiness based Collusive Q&A Spam Detection method (PBCSD) to identify the deceptive questions and answers. Specifically, we intensively study the campaign process of crowdsourcing tasks and summarize the clues in the Q&As’ vocabulary usage level when collusive attacks are launched. Based on the clues, we extract the Q&A groups using frequent pattern mining and further purify them by the burstiness on posting time of Q&As. By designing several discriminative features at the Q&A group level, multiple machine learning based classifiers can be used to judge the groups as deceptive or ordinary, and the Q&As in deceptive groups are finally identified as collusive Q&A spam. We evaluate the proposed PBCSD method in a real-world dataset collected from Baidu Zhidao, a famous CQA platform in China, and the experimental results demonstrate the PBCSD is effective for collusive Q&A spam detection and outperforms a number of state-of-art methods.

State-of-the-art approaches in gait analysis usually rely on one isolated tracking system, generating insufficient data for complex use cases such as sports, rehabilitation, and MedTech. We address the opportunity to comprehensively understand human motion by a novel data model combining several motion-tracking methods. The model aggregates pose estimation by captured videos and EMG and EIT sensor data synchronously to gain insights into muscle activities. Our demonstration with biceps curl and sitting/standing pose generates time-synchronous data and delivers insights into our experiment’s usability, advantages, and challenges.

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.