Biblio

Conventional adversarial defenses reduce classification accuracy whether or not a model is under attacks. Moreover, most of image processing based defenses are defeated due to the problem of obfuscated gradients. In this paper, we propose a new adversarial defense which is a defensive transform for both training and test images inspired by perceptual image encryption methods. The proposed method utilizes a block-wise pixel shuffling method with a secret key. The experiments are carried out on both adaptive and non-adaptive maximum-norm bounded white-box attacks while considering obfuscated gradients. The results show that the proposed defense achieves high accuracy (91.55%) on clean images and (89.66%) on adversarial examples with noise distance of 8/255 on CFAR-10 dataset. Thus, the proposed defense outperforms state-of-the-art adversarial defenses including latent adversarial training, adversarial training and thermometer encoding.

Federated learning (FL) is the rapidly developing machine learning technique that is used to perform collaborative model training over decentralized datasets. FL enables privacy-preserving model development whereby the datasets are scattered over a large set of data producers (i.e., devices and/or systems). These data producers train the learning models, encapsulate the model updates with differential privacy techniques, and share them to centralized systems for global aggregation. However, these centralized models are always prone to adversarial attacks (such as data-poisoning and model poisoning attacks) due to a large number of data producers. Hence, FL methods need to ensure fairness and high-quality model availability across all the participants in the underlying AI systems. In this paper, we propose a novel FL framework, called FairFed, to meet fairness and high-quality data requirements. The FairFed provides a fairness mechanism to detect adversaries across the devices and datasets in the FL network and reject their model updates. We use a Python-simulated FL framework to enable large-scale training over MNIST dataset. We simulate a cross-device model training settings to detect adversaries in the training network. We used TensorFlow Federated and Python to implement the fairness protocol, the deep neural network, and the outlier detection algorithm. We thoroughly test the proposed FairFed framework with random and uniform data distributions across the training network and compare our initial results with the baseline fairness scheme. Our proposed work shows promising results in terms of model accuracy and loss.

Generative Adversarial Network (GAN) and its variants have shown promising results in generating synthetic data. However, the issues with GANs are: (i) the learning happens around the training samples and the model often ends up remembering them, consequently, compromising the privacy of individual samples - this becomes a major concern when GANs are applied to training data including personally identifiable information, (ii) the randomness in generated data - there is no control over the specificity of generated samples. To address these issues, we propose imdpGAN-an information maximizing differentially private Generative Adversarial Network. It is an end-to-end framework that simultaneously achieves privacy protection and learns latent representations. With experiments on MNIST dataset, we show that imdpGAN preserves the privacy of the individual data point, and learns latent codes to control the specificity of the generated samples. We perform binary classification on digit pairs to show the utility versus privacy trade-off. The classification accuracy decreases as we increase privacy levels in the framework. We also experimentally show that the training process of imdpGAN is stable but experience a 10-fold time increase as compared with other GAN frameworks. Finally, we extend imdpGAN framework to CelebA dataset to show how the privacy and learned representations can be used to control the specificity of the output.

In recent years, we have faced a series of online threats. The continuous malicious attacks on the network have directly caused a huge threat to the user's spirit and property. In order to deal with the complex security situation in today's network environment, an intelligent network situational awareness model based on deep neural networks is proposed. Use the nonlinear characteristics of the deep neural network to solve the nonlinear fitting problem, establish a network security situation assessment system, take the situation indicators output by the situation assessment system as a guide, and collect on the main data features according to the characteristics of the network attack method, the main data features are collected and the data is preprocessed. This model designs and trains a 4-layer neural network model, and then use the trained deep neural network model to understand and analyze the network situation data, so as to build the network situation perception model based on deep neural network. The deep neural network situational awareness model designed in this paper is used as a network situational awareness simulation attack prediction experiment. At the same time, it is compared with the perception model using gray theory and Support Vector Machine(SVM). The experiments show that this model can make perception according to the changes of state characteristics of network situation data, establish understanding through learning, and finally achieve accurate prediction of network attacks. Through comparison experiments, datatypized neural network deep neural network situation perception model is proved to be effective, accurate and superior.

With the increasing popularity of the Internet, mobile Internet and the Internet of Things, the current network environment continues to become more complicated. Due to the increasing variety and severity of cybersecurity threats, traditional means of network security protection have ushered in a huge challenge. The network security posture prediction can effectively predict the network development trend in the future time based on the collected network history data, so this paper proposes an algorithm based on simulated annealing-particle swarm algorithm to optimize improved Elman neural network parameters to achieve posture prediction for network security. Taking advantage of the characteristic that the value of network security posture has periodicity, a simulated annealing algorithm is introduced along with an improved particle swarm algorithm to solve the problem that neural network training is prone to fall into a local optimal solution and achieve accurate prediction of the network security posture. Comparison of the proposed scheme with existing prediction methods validates that the scheme has a good posture prediction accuracy.

Protecting and preventing sensitive data from being used inappropriately has become a challenging task. Even a small mistake in securing data can be exploited by phishing attacks to release private information such as passwords or financial information to a malicious actor. Phishing has now proven so successful, it is the number one attack vector. Many approaches have been proposed to protect against this type of cyber-attack, from additional staff training, enriched spam filters to large collaborative databases of known threats such as PhishTank and OpenPhish. However, they mostly rely upon a user falling victim to an attack and manually adding this new threat to the shared pool, which presents a constant disadvantage in the fight back against phishing. In this paper, we propose a novel approach to protect against phishing attacks using binary visualisation and machine learning. Unlike previous work in this field, our approach uses an automated detection process and requires no further user interaction, which allows faster and more accurate detection process. The experiment results show that our approach has high detection rate.

Quantized deep neural networks (QDNNs) are necessary for low-power, high throughput, and embedded applications. Previous studies mostly focused on developing optimization methods for the quantization of given models. However, quantization sensitivity depends on the model architecture. Also, the characteristics of weight and activation quantization are quite different. This study proposes a holistic approach for the optimization of QDNNs, which contains QDNN training methods as well as quantization-friendly architecture design. Synthesized data is used to visualize the effects of weight and activation quantization. The results indicate that deeper models are more prone to activation quantization, while wider models improve the resiliency to both weight and activation quantization.

Cloud computing is applied in various domains, among which social recommender systems are well-received because of their effectivity to provide suggestions for users. Social recommender systems perform well in alleviating cold start problem, but it suffers from shilling attack due to its natural openness. Shilling attack is an injection attack mainly acting on the training process of machine learning, which aims to advance or suppress the recommendation ranking of target items. Some researchers have studied the influence of shilling attacks in two perspectives simultaneously, which are user-item's rating and user-user's relation. However, they take more consideration into user-item's rating, and up to now, the construction of user-user's relation has not been explored in depth. To explore shilling attacks with complex relations, in this paper, we propose two novel attack models based on triangle relations in social networks. Furthermore, we explore the influence of these models on five social recommendation algorithms. The experimental results on three datasets show that the recommendation can be affected by the triangle relation attacks. The attack model combined with triangle relation has a better attack effect than the model only based on rating injection and the model combined with random relation. Besides, we compare the functions of triangle relations in friend recommendation and product recommendation.

This paper investigates the effects of real time visual biofeedback for improving sports performance using a large scale immersive mixed reality system in which users are able to play a simulated game of curling. The users slide custom curling stones across the floor onto a projected target whose size is dictated by the user’s stress-related physiological measure; heart rate (HR). The higher HR the player has, the smaller the target will be, and vice-versa. In the experiment participants were asked to compete in three different conditions: baseline, with and without the proposed biofeedback. The results show that when providing a visual representation of the player’s HR or "choking" in competition, it helped the player understand their condition and improve competition performance (P-value of 0.0391).

In recent times, an increasing amount of intelligent electronic devices (IEDs) are being deployed to make power systems more reliable and economical. While these technologies are necessary for realizing a cyber-physical infrastructure for future smart power grids, they also introduce new vulnerabilities in the grid to different cyber-attacks. Traditional methods such as state vector estimation (SVE) are not capable of identifying cyber-attacks while the geometric information is also injected as an attack vector. In this paper, a machine learning based smart grid attack identification method is proposed. The proposed method is carried out by first collecting smart grid power flow data for machine learning training purposes which is later used to classify the attacks. The performance of both the proposed SVM method and the traditional SVE method are validated on IEEE 14, 30, 39, 57 and 118 bus systems, and the performance regarding the scale of the power system is evaluated. The results show that the SVM-based method performs better than the SVE-based in attack identification over a much wider scale of power systems.

Quasi-Realistic cyber-physical system (QR-CPS) testbed architecture and operational environment are critical for testing and validating various cyber attack-defense algorithms for the wide-area resilient power systems. These QR-CPS testbed environments provide a realistic platform for conducting the Grid Exercise (GridEx), CPS security training, and attack-defense exercise at a broader scale for the cybersecurity of Energy Delivery Systems. The NERC has established a tabletop based GridEx platform for the North American power utilities to demonstrate how they would respond to and recover from cyber threats and incidents. The NERC-GridEx is a bi-annual activity with tabletop attack injects and incidence response management. There is a significant need to build a testbed-based hands-on GridEx for the utilities by leveraging the CPS testbeds, which imitates the pragmatic CPS grid environment. We propose a CPS testbed-based Quasi-Realistic Grid Exercise (QR-GridEx), which is a model after the NERC's tabletop GridEx. We have designed the CPS testbed-based QR-GridEx into two parts. Part-I focuses on the modeling of synthetic grid models for the utilities, including SCADA and WAMS communications, and attack-and-defense software systems; and the Part-II focuses on the incident response management and risk-based CPS grid investment strategies. This paper presents the Part-I of the CPS testbed-based QRGridEx, which includes modeling of the synthetic grid models in the real-time digital simulator, stealthy, and coordinated cyberattack vectors, and integration of intrusion/anomaly detection systems. We have used our existing HIL CPS security testbed to demonstrate the testbed-based QR-GridEx for a Texas-2000 bus US synthetic grid model and the IEEE-39 bus grid models. The experiments demonstrated significant results by 100% real-time performance with zero overruns for grid impact characteristics against stealthy and coordinated cyberattack vectors.

By the multi-layer nonlinear mapping and the semantic feature extraction of the deep learning, a deep learning network is proposed for video face detection to overcome the challenge of detecting faces rapidly and accurately in video with changeable background. Particularly, a pre-training procedure is used to initialize the network parameters to avoid falling into the local optimum, and the greedy layer-wise learning is introduced in the pre-training to avoid the training error transfer in layers. Key to the network is that the probability of neurons models the status of human brain neurons which is a continuous distribution from the most active to the least active and the hidden layer’s neuron number decreases layer-by-layer to reduce the redundant information of the input data. Moreover, the skin color detection is used to accelerate the detection speed by generating candidate regions. Experimental results show that, besides the faster detection speed and robustness against face rotation, the proposed method possesses lower false detection rate and lower missing detection rate than traditional algorithms.

The security of energy supply in a power grid critically depends on the ability to accurately estimate the state of the system. However, manipulated power flow measurements can potentially hide overloads and bypass the bad data detection scheme to interfere the validity of estimated states. In this paper, we use an autoencoder neural network to detect anomalous system states and investigate the impact of hyperparameters on the detection performance for false data injection attacks that target power flows. Experimental results on the IEEE 118 bus system indicate that the proposed mechanism has the ability to achieve satisfactory learning efficiency and detection accuracy.

The Global Positioning System (GPS) can determine the position of any person or object on earth based on satellite signals. But when inside the building, the GPS cannot receive signals, the indoor positioning system will determine the precise position. How to achieve more precise positioning is the difficulty of an indoor positioning system now. In this paper, we proposed an ultra-wideband fingerprinting positioning method based on a convolutional neural network (CNN), and we collect the dataset in a room to test the model, then compare our method with the existing method. In the experiment, our method can reach an accuracy of 98.36%. Compared with other fingerprint positioning methods our method has a great improvement in robustness. That results show that our method has good practicality while achieves higher accuracy.

Learning-based approaches often outperform hand-coded algorithmic solutions for many problems in robotics. However, learning long-horizon tasks on real robot hardware can be intractable, and transferring a learned policy from simulation to reality is still extremely challenging. We present a novel approach to model-free reinforcement learning that can leverage existing sub-optimal solutions as an algorithmic prior during training and deployment. During training, our gated fusion approach enables the prior to guide the initial stages of exploration, increasing sample-efficiency and enabling learning from sparse long-horizon reward signals. Importantly, the policy can learn to improve beyond the performance of the sub-optimal prior since the prior's influence is annealed gradually. During deployment, the policy's uncertainty provides a reliable strategy for transferring a simulation-trained policy to the real world by falling back to the prior controller in uncertain states. We show the efficacy of our Multiplicative Controller Fusion approach on the task of robot navigation and demonstrate safe transfer from simulation to the real world without any fine-tuning. The code for this project is made publicly available at https://sites.google.com/view/mcf-nav/home.

Photo- and video-realistic generation techniques have become a reality following the advent of deep neural networks. Consequently, there are immense concerns regarding the difficulty in differentiating what content is real from what is synthetic. An example of video-realistic generation techniques is the infamous Deepfakes, which exploit the main modality by which humans identify each other. Deepfakes are a category of synthetic face generation methods and are commonly based on generative adversarial networks. In this article, we propose a novel two-step synthetic face image detection method in which general-purpose features are extracted in a first step, trivializing the task of detecting synthetic images. The anomaly detector predicts the conditional probabilities for observing every individual pixel in the image and is trained on pristine data only. The extracted anomaly features demonstrate true generalization capacity across widely different unknown synthesis methods while showing a minimal loss in performance with regard to the detection of known synthetic samples.

Hyperbox classifier is an efficient tool for modern pattern classification problems due to its transparency and rigorous use of Euclidian geometry. Fuzzy min-max (FMM) network efficiently implements the hyperbox classifier, and has been modified several times to yield better classification accuracy. However, the obtained accuracy is not up to the mark. Therefore, in this paper, a new improved FMM (IFMM) network is proposed to increase the accuracy rate. In the proposed IFMM network, a modified constraint is employed to check the expandability of a hyperbox. It also uses semiperimeter of the hyperbox along with k-nearest mechanism to select the expandable hyperbox. In the proposed IFMM, the contraction rules of conventional FMM and enhanced FMM (EFMM) are also modified using semiperimeter of a hyperbox in order to balance the size of both overlapped hyperboxes. Experimental results show that the proposed IFMM network outperforms the FMM, k-nearest FMM, and EFMM by yielding more accuracy rate with less number of hyperboxes. The proposed methods are also applied to histopathological images to know the best magnification factor for classification.

Existing Lipschitz-based provable defences to adversarial examples only cover the L2 threat model. We introduce the first bound that makes use of Lipschitz continuity to provide a more general guarantee for threat models based on any Lp norm. Additionally, a new strategy is proposed for designing network architectures that exhibit superior provable adversarial robustness over conventional convolutional neural networks. Experiments are conducted to validate our theoretical contributions, show that the assumptions made during the design of our novel architecture hold in practice, and quantify the empirical robustness of several Lipschitz-based adversarial defence methods.

The new switching architecture of the neural networks was proposed. The switching neural networks consist of the neurons and the switchers. The goal is to reduce expenses on the artificial neural network design and training. For realization of complex models, algorithms and methods of management the neural networks of the big size are required. The number of the interconnection links “everyone with everyone” grows with the number of neurons. The training of big neural networks requires the resources of supercomputers. Time of training of neural networks also depends on the number of neurons in the network. Switching neural networks are divided into fragments connected by the switchers. Training of switcher neuron network is provided by fragments. On the basis of switching neural networks the devices of associative memory were designed with the number of neurons comparable to the human brain.

Steganalysis is an interesting classification problem in order to discriminate the images, including hidden messages from the clean ones. There are many methods, including deep CNN networks to extract fine features for this classification task. Nevertheless, a few researches have been conducted to improve the final classifier. Some state-of-the-art methods try to ensemble the networks by a voting strategy to achieve more stable performance. In this paper, a selection phase is proposed to filter improper networks before any voting. This filtering is done by a binary relevance multi-label classification approach. The Logistic Regression (LR) is chosen here as the last layer of network for classification. The large-margin Fisher’s linear discriminant (FLD) classifier is assigned to each one of the networks. It learns to discriminate the training instances which associated network is suitable for or not. Xu-Net, one of the most famous state-of-the-art Steganalysis models, is chosen as the base networks. The proposed method with different approaches is applied on the BOSSbase dataset and is compared with traditional voting and also some state-of-the-art related ensemble techniques. The results show significant accuracy improvement of the proposed method in comparison with others.

Static analysis tools are programs that run on source code prior to their compilation to binary executables and attempt to find flaws or defects in the code during the early stages of development. If left unresolved, these flaws could pose security risks. While numerous static analysis tools exist, there is no single tool that is optimal. Therefore, many static analysis tools are often used to analyze code. Further, some of the alerts generated by the static analysis tools are low-priority or false alarms. Machine learning algorithms have been developed to distinguish between true alerts and false alarms, however significant man hours need to be dedicated to labeling data sets for training. This study investigates the use of active learning to reduce the number of labeled alerts needed to adequately train a classifier. The numerical experiments demonstrate that a query by committee active learning algorithm can be utilized to significantly reduce the number of labeled alerts needed to achieve similar performance as a classifier trained on a data set of nearly 60,000 labeled alerts.

Protecting sensitive information from illegal access and misuse is crucial to all organizations. An inappropriate Information Security (IS) policy and procedures are not only a suitable environment for an outsider attack but also a good chance for the insiders' misuse. In this paper, we will discuss the roles of an organization in information security and how human behavior affects the Information Security System (ISS). How an organization can create and instill an effective information security culture in an organization to improve their information safeguards. The findings in this review can be used to further researches and will be useful for organizations to improve their information security structure (ISC).

Every public or private area today is preferred to be under surveillance to ensure high levels of security. Since the surveillance happens round the clock, data gathered as a result is huge and requires a lot of manual work to go through every second of the recorded videos. This paper presents a system which can detect anomalous behaviors and alarm the user on the type of anomalous behavior. Since there are a myriad of anomalies, the classification of anomalies had to be narrowed down. There are certain anomalies which are generally seen and have a huge impact on public safety, such as explosions, road accidents, assault, shooting, etc. To narrow down the variations, this system can detect explosion, road accidents, shooting, and fighting and even output the frame of their occurrence. The model has been trained with videos belonging to these classes. The dataset used is UCF Crime dataset. Learning patterns from videos requires the learning of both spatial and temporal features. Convolutional Neural Networks (CNN) extract spatial features and Long Short-Term Memory (LSTM) networks learn the sequences. The classification, using an CNN-LSTM model achieves an accuracy of 85%.

In this paper, it first analyzed the characteristics of Advanced Persistent Threat (APT). according to APT attack, this paper established an BP neural network optimized by improved adaptive genetic algorithm to predict the security risk of nodes in the network. and calculated the path of APT attacks with the maximum possible attack. Finally, experiments verify the effectiveness and correctness of the algorithm by simulating attacks. Experiments show that this model can effectively evaluate the security situation in the network, For the defenders to adopt effective measures defend against APT attacks, thus improving the security of the network.

Natural language processing (NLP) have been recently used to extract clinical information from free text in Electronic Health Record (EHR). In clinical NLP one challenge is that the meaning of clinical entities is heavily affected by assertion modifiers such as negation, uncertain, hypothetical, experiencer and so on. Incorrect assertion assignment could cause inaccurate diagnosis of patients' condition or negatively influence following study like disease modeling. Thus, clinical NLP systems which can detect assertion status of given target medical findings (e.g. disease, symptom) in clinical context are highly demanded. Here in this work, we propose a deep-learning system based on word embedding, RNN and attention mechanism (more specifically: Attention-based Bidirectional Long Short-Term Memory networks) for assertion detection in clinical notes. Unlike previous state-of-art methods which require knowledge input or feature engineering, our system is a knowledge poor machine learning system and can be easily extended or transferred to other domains. The evaluation of our system on public benchmarking corpora demonstrates that a knowledge poor deep-learning system can also achieve high performance for detecting negation and assertions comparing to state-of-the-art systems.