Biblio

Wide adoption of artificial neural networks in various domains has led to an increasing interest in defending adversarial attacks against them. Preprocessing defense methods such as pixel discretization are particularly attractive in practice due to their simplicity, low computational overhead, and applicability to various systems. It is observed that such methods work well on simple datasets like MNIST, but break on more complicated ones like ImageNet under recently proposed strong white-box attacks. To understand the conditions for success and potentials for improvement, we study the pixel discretization defense method, including more sophisticated variants that take into account the properties of the dataset being discretized. Our results again show poor resistance against the strong attacks. We analyze our results in a theoretical framework and offer strong evidence that pixel discretization is unlikely to work on all but the simplest of the datasets. Furthermore, our arguments present insights why some other preprocessing defenses may be insecure.

In this paper, we consider one of the approaches to the study of the characteristics of an information system that is under the influence of various factors, and their management using neural networks and wavelet transforms based on determining the relationship between the modified state of the information system and the possibility of dynamic analysis of effects. At the same time, the process of influencing the information system includes the following components: impact on the components providing the functions of the information system; determination of the result of exposure; analysis of the result of exposure; response to the result of exposure. As an input signal, the characteristics of the means that affect are taken. The system includes an adaptive response unit, the input of which receives signals about the prerequisites for changes, and at the output, this unit generates signals for the inclusion of appropriate means to eliminate or compensate for these prerequisites or directly the changes in the information system.

Video Steganography is an extension of image steganography where any kind of file in any extension is hidden into a digital video. The video content is dynamic in nature and this makes the detection of hidden data difficult than other steganographic techniques. The main motive of using video steganography is that the videos can store large amount of data in it. This paper focuses on security using the combination of hybrid neural networks and hash function for determining the best bits in the cover video to embed the secret data. For the embedding process, the cover video and the data to be hidden is uploaded. Then the hash algorithm and neural networks are applied to form the stego video. For the extraction process, the reverse process is applied and the secret data is obtained. All experiments are done using MatLab2016a software.

Spams are unsolicited and unnecessary messages which may contain harmful codes or links for activation of malicious viruses and spywares. Increasing popularity of social networks attracts the spammers to perform malicious activities in social networks. So an efficient spam detection method is necessary for social networks. In this paper, feed forward neural network with back propagation based spam detection model is proposed. The quality of the learning process is improved by tuning initial weights of feed forward neural network using proposed enhanced step size firefly algorithm which reduces the time for finding optimal weights during the learning process. The model is applied for twitter dataset and the experimental results show that, the proposed model performs well in terms of accuracy and detection rate and has lower false positive rate. 

Machine learning systems have had enormous success in a wide range of fields from computer vision, natural language processing, and anomaly detection. However, such systems are vulnerable to attackers who can cause deliberate misclassification by introducing small perturbations. With machine learning systems being proposed for cyber attack detection such attackers are cause for serious concern. Despite this the vast majority of adversarial machine learning security research is focused on the image domain. This work gives a brief overview of adversarial machine learning and machine learning used in cyber attack detection and suggests key differences between the traditional image domain of adversarial machine learning and the cyber domain. Finally we show an adversarial machine learning attack on an industrial control system.

Recent work proposed the concept of backdoor attacks on deep neural networks (DNNs), where misclassification rules are hidden inside normal models, only to be triggered by very specific inputs. However, these "traditional" backdoors assume a context where users train their own models from scratch, which rarely occurs in practice. Instead, users typically customize "Teacher" models already pretrained by providers like Google, through a process called transfer learning. This customization process introduces significant changes to models and disrupts hidden backdoors, greatly reducing the actual impact of backdoors in practice. In this paper, we describe latent backdoors, a more powerful and stealthy variant of backdoor attacks that functions under transfer learning. Latent backdoors are incomplete backdoors embedded into a "Teacher" model, and automatically inherited by multiple "Student" models through transfer learning. If any Student models include the label targeted by the backdoor, then its customization process completes the backdoor and makes it active. We show that latent backdoors can be quite effective in a variety of application contexts, and validate its practicality through real-world attacks against traffic sign recognition, iris identification of volunteers, and facial recognition of public figures (politicians). Finally, we evaluate 4 potential defenses, and find that only one is effective in disrupting latent backdoors, but might incur a cost in classification accuracy as tradeoff.

Recent advances in networking and communication technologies have enabled Internet-of-Things (IoT) devices to communicate more frequently and faster. An IoT device typically transmits data over the Internet which is an insecure channel. Cyber attacks such as denial-of-service (DoS), man-in-middle, and SQL injection are considered as big threats to IoT devices. In this paper, an anomaly-based intrusion detection scheme is proposed that can protect sensitive information and detect novel cyber-attacks. The Artificial Bee Colony (ABC) algorithm is used to train the Random Neural Network (RNN) based system (RNN-ABC). The proposed scheme is trained on NSL-KDD Train+ and tested for unseen data. The experimental results suggest that swarm intelligence and RNN successfully classify novel attacks with an accuracy of 91.65%. Additionally, the performance of the proposed scheme is also compared with a hybrid multilayer perceptron (MLP) based intrusion detection system using sensitivity, mean of mean squared error (MMSE), the standard deviation of MSE (SDMSE), best mean squared error (BMSE) and worst mean squared error (WMSE) parameters. All experimental tests confirm the robustness and high accuracy of the proposed scheme.

Smart mobility management would be an important prerequisite for future fog computing systems. In this research, we propose a learning-based handover optimization for the Internet of Vehicles that would assist the smooth transition of device connections and offloaded tasks between fog nodes. To accomplish this, we make use of machine learning algorithms to learn from vehicle interactions with fog nodes. Our approach uses a three-layer feed-forward neural network to predict the correct fog node at a given location and time with 99.2 % accuracy on a test set. We also implement a dual stacked recurrent neural network (RNN) with long short-term memory (LSTM) cells capable of learning the latency, or cost, associated with these service requests. We create a simulation in JAMScript using a dataset of real-world vehicle movements to create a dataset to train these networks. We further propose the use of this predictive system in a smarter request routing mechanism to minimize the service interruption during handovers between fog nodes and to anticipate areas of low coverage through a series of experiments and test the models' performance on a test set.

The greatest threat towards securing the organization and its assets are no longer the attackers attacking beyond the network walls of the organization but the insiders present within the organization with malicious intent. Existing approaches helps to monitor, detect and prevent any malicious activities within an organization's network while ignoring the human behavior impact on security. In this paper we have focused on user behavior profiling approach to monitor and analyze user behavior action sequence to detect insider threats. We present an ensemble hybrid machine learning approach using Multi State Long Short Term Memory (MSLSTM) and Convolution Neural Networks (CNN) based time series anomaly detection to detect the additive outliers in the behavior patterns based on their spatial-temporal behavior features. We find that using Multistate LSTM is better than basic single state LSTM. The proposed method with Multistate LSTM can successfully detect the insider threats providing the AUC of 0.9042 on train data and AUC of 0.9047 on test data when trained with publically available dataset for insider threats.

In this paper, we propose a method for improving the robustness of real-time facial expression recognition. Although there are many ways to improve the accuracy of facial expression recognition, a revamp of the training framework and image preprocessing allow better results in applications. One existing problem is that when the camera is capturing images in high speed, changes in image characteristics may occur at certain moments due to the influence of light and other factors. Such changes can result in incorrect recognition of the human facial expression. To solve this problem for smooth system operation and maintenance of recognition speed, we take changes in image characteristics at high speed capturing into account. The proposed method does not use the immediate output for reference, but refers to the previous image for averaging to facilitate recognition. In this way, we are able to reduce interference by the characteristics of the images. The experimental results show that after adopting this method, overall robustness and accuracy of facial expression recognition have been greatly improved compared to those obtained by only the convolution neural network (CNN).

Advances in our understanding of the nature of cognition in its myriad forms (Embodied, Embedded, Extended, and Enactive) displayed in all living beings (cellular organisms, animals, plants, and humans) and new theories of information, info-computation and knowledge are throwing light on how we should build software systems in the digital universe which mimic and interact with intelligent, sentient and resilient beings in the physical universe. Recent attempts to infuse cognition into computing systems to push the boundaries of Church-Turing thesis have led to new computing models that mimic biological systems in encoding knowledge structures using both algorithms executed in stored program control machines and neural networks. This paper presents a new model and implements an application as hierarchical named network composed of microservices to create a managed process workflow by enabling dynamic configuration and reconfiguration of the microservice network. We demonstrate the resiliency, efficiency and scaling of the named microservice network using a novel edge cloud platform by Platina Systems. The platform eliminates the need for Virtual Machine overlay and provides high performance and low-latency with L3 based 100 GbE network and SSD support with RDMA and NVMeoE. The hierarchical named microservice network using Kubernetes provisioning stack provides all the cloud features such as elasticity, autoscaling, self-repair and live-migration without reboot. The model is derived from a recent theoretical framework for unification of different models of computation using "Structural Machines.'' They are shown to simulate Turing machines, inductive Turing machines and also are proved to be more efficient than Turing machines. The structural machine framework with a hierarchy of controllers managing the named service connections provides dynamic reconfiguration of the service network from browsers to database to address rapid fluctuations in the demand for or the availability of resources without having to reconfigure IP address base networks.

Hiring seasonal workers in call centers to provide customer service is a common practice in B2C companies. The quality of service delivered by both contracting and employee customer service agents depends heavily on the domain knowledge available to them. When observing the internal group messaging channels used by agents, we found that similar questions are often asked repetitively by different agents, especially from less experienced ones. The goal of our work is to leverage the promising advances in conversational AI to provide a chatbot-like mechanism for assisting agents in promptly resolving a customer's issue. In this paper, we develop a neural-based conversational solution that employs BiLSTM with attention mechanism and demonstrate how our system boosts the effectiveness of customer support agents. In addition, we discuss the design principles and the necessary considerations for our system. We then demonstrate how our system, named "Isa" (Intuit Smart Agent), can help customer service agents provide a high-quality customer experience by reducing customer wait time and by applying the knowledge accumulated from customer interactions in future applications.

We investigate a method of building a closed domain intelligent conversational agent using deep neural networks. A conversational agent is a dialog system intended to converse with a human, with a coherent structure. Our conversational agent uses a retrieval based model that identifies the intent of the input user query and maps it to a knowledge base to return appropriate results. Human conversations are based on context, but existing conversational agents are context insensitive. To overcome this limitation, our system uses a simple stack based context identification and storage system. The conversational agent generates responses according to the current context of conversation. allowing more human-like conversations.

This paper investigates the use of deep reinforcement learning (DRL) in the design of a "universal" MAC protocol referred to as Deep-reinforcement Learning Multiple Access (DLMA). The design framework is partially inspired by the vision of DARPA SC2, a 3-year competition whereby competitors are to come up with a clean-slate design that "best share spectrum with any network(s), in any environment, without prior knowledge, leveraging on machine-learning technique". While the scope of DARPA SC2 is broad and involves the redesign of PHY, MAC, and Network layers, this paper's focus is narrower and only involves the MAC design. In particular, we consider the problem of sharing time slots among a multiple of time-slotted networks that adopt different MAC protocols. One of the MAC protocols is DLMA. The other two are TDMA and ALOHA. The DRL agents of DLMA do not know that the other two MAC protocols are TDMA and ALOHA. Yet, by a series of observations of the environment, its own actions, and the rewards - in accordance with the DRL algorithmic framework - a DRL agent can learn the optimal MAC strategy for harmonious co-existence with TDMA and ALOHA nodes. In particular, the use of neural networks in DRL (as opposed to traditional reinforcement learning) allows for fast convergence to optimal solutions and robustness against perturbation in hyper- parameter settings, two essential properties for practical deployment of DLMA in real wireless networks.

In last decades, the web and online services have revolutionized the modern world. However, by increasing our dependence on online services, as a result, online security threats are also increasing rapidly. One of the most common online security threats is a so-called Phishing attack, the purpose of which is to mimic a legitimate website such as online banking, e-commerce or social networking website in order to obtain sensitive data such as user-names, passwords, financial and health-related information from potential victims. The problem of detecting phishing websites has been addressed many times using various methodologies from conventional classifiers to more complex hybrid methods. Recent advancements in deep learning approaches suggested that the classification of phishing websites using deep learning neural networks should outperform the traditional machine learning algorithms. However, the results of utilizing deep neural networks heavily depend on the setting of different learning parameters. In this paper, we propose a swarm intelligence based approach to parameter setting of deep learning neural network. By applying the proposed approach to the classification of phishing websites, we were able to improve their detection when compared to existing algorithms.

Recent insights on language and vision with neural networks have been successfully applied to simple single-image visual question answering. However, to tackle real-life question answering problems on multimedia collections such as personal photos, we have to look at whole collections with sequences of photos or videos. When answering questions from a large collection, a natural problem is to identify snippets to support the answer. In this paper, we describe a novel neural network called Focal Visual-Text Attention network (FVTA) for collective reasoning in visual question answering, where both visual and text sequence information such as images and text metadata are presented. FVTA introduces an end-to-end approach that makes use of a hierarchical process to dynamically determine what media and what time to focus on in the sequential data to answer the question. FVTA can not only answer the questions well but also provides the justifications which the system results are based upon to get the answers. FVTA achieves state-of-the-art performance on the MemexQA dataset and competitive results on the MovieQA dataset.

The localization and classification of cryptographic functions in binary files is a growing challenge in information security, not least because of the increasing use of such functions in malware. Nevertheless, it is still a time consuming and laborious task. Some of the most commonly used techniques are based on dynamic methods, signatures or manual reverse engineering. In this paper we present FALKE-MC, a novel framework that creates classifiers for arbitrary cryptographic algorithms from sample binaries. It processes multiple file formats and architectures and is easily expandable due to its modular design. Functions are automatically detected and features as well as constants are extracted. They are used to train a neural network, which can then be applied to classify functions in unknown binary files. The framework is fully automated, from the input of binary files and the creation of a classifier through to the output of classification results. In addition to that, it can deal with class imbalance between cryptographic and non-cryptographic samples during training. Our evaluation shows that this approach offers a high detection rate in combination with a low false positive rate. We are confident that FALKE-MC can accelerate the localization and classification of cryptographic functions in practice.

Recently researchers have proposed using deep learning-based systems for malware detection. Unfortunately, all deep learning classification systems are vulnerable to adversarial learning-based attacks, or adversarial attacks, where miscreants can avoid detection by the classification algorithm with very few perturbations of the input data. Previous work has studied adversarial attacks against static analysis-based malware classifiers which only classify the content of the unknown file without execution. However, since the majority of malware is either packed or encrypted, malware classification based on static analysis often fails to detect these types of files. To overcome this limitation, anti-malware companies typically perform dynamic analysis by emulating each file in the anti-malware engine or performing in-depth scanning in a virtual machine. These strategies allow the analysis of the malware after unpacking or decryption. In this work, we study different strategies of crafting adversarial samples for dynamic analysis. These strategies operate on sparse, binary inputs in contrast to continuous inputs such as pixels in images. We then study the effects of two, previously proposed defensive mechanisms against crafted adversarial samples including the distillation and ensemble defenses. We also propose and evaluate the weight decay defense. Experiments show that with these three defenses, the number of successfully crafted adversarial samples is reduced compared to an unprotected baseline system. In particular, the ensemble defense is the most resilient to adversarial attacks. Importantly, none of the defenses significantly reduce the classification accuracy for detecting malware. Finally, we show that while adding additional hidden layers to neural models does not significantly improve the malware classification accuracy, it does significantly increase the classifier's robustness to adversarial attacks.

Early detection of new kinds of malware always plays an important role in defending the network systems. Especially, if intelligent protection systems could themselves detect an existence of new malware types in their system, even with a very small number of malware samples, it must be a huge benefit for the organization as well as the social since it help preventing the spreading of that kind of malware. To deal with learning from few samples, term ``one-shot learning'' or ``fewshot learning'' was introduced, and mostly used in computer vision to recognize images, handwriting, etc. An approach introduced in this paper takes advantage of One-shot learning algorithms in solving the malware classification problem by using Memory Augmented Neural Network in combination with malware's API calls sequence, which is a very valuable source of information for identifying malware behavior. In addition, it also use some advantages of the development in Natural Language Processing field such as word2vec, etc. to convert those API sequences to numeric vectors before feeding to the one-shot learning network. The results confirm very good accuracies compared to the other traditional methods.

Better understanding of mobile applications' behaviors would lead to better malware detection/classification and better app recommendation for users. In this work, we design a framework AppDNA to automatically generate a compact representation for each app to comprehensively profile its behaviors. The behavior difference between two apps can be measured by the distance between their representations. As a result, the versatile representation can be generated once for each app, and then be used for a wide variety of objectives, including malware detection, app categorizing, plagiarism detection, etc. Based on a systematic and deep understanding of an app's behavior, we propose to perform a function-call-graph-based app profiling. We carefully design a graph-encoding method to convert a typically extremely large call-graph to a 64-dimension fix-size vector to achieve robust app profiling. Our extensive evaluations based on 86,332 benign and malicious apps demonstrate that our system performs app profiling (thus malware detection, classification, and app recommendation) to a high accuracy with extremely low computation cost: it classifies 4024 (benign/malware) apps using around 5.06 second with accuracy about 93.07%; it classifies 570 malware's family (total 21 families) using around 0.83 second with accuracy 82.3%; it classifies 9,730 apps' functionality with accuracy 33.3% for a total of 7 categories and accuracy of 88.1 % for 2 categories.

Modern industrial control systems (ICS) act as victims of cyber attacks more often in last years. These attacks are hard to detect and their consequences can be catastrophic. Cyber attacks can cause anomalies in the work of the ICS and its technological equipment. The presence of mutual interference and noises in this equipment significantly complicates anomaly detection. Moreover, the traditional means of protection, which used in corporate solutions, require updating with each change in the structure of the industrial process. An approach based on the machine learning for anomaly detection was used to overcome these problems. It complements traditional methods and allows one to detect signal correlations and use them for anomaly detection. Additional Tennessee Eastman Process Simulation Data for Anomaly Detection Evaluation dataset was analyzed as example of industrial process. In the course of the research, correlations between the signals of the sensors were detected and preliminary data processing was carried out. Algorithms from the most common techniques of machine learning (decision trees, linear algorithms, support vector machines) and deep learning models (neural networks) were investigated for industrial process anomaly detection task. It's shown that linear algorithms are least demanding on computational resources, but they don't achieve an acceptable result and allow a significant number of errors. Decision tree-based algorithms provided an acceptable accuracy, but the amount of RAM, required for their operations, relates polynomially with the training sample volume. The deep neural networks provided the greatest accuracy, but they require considerable computing power for internal calculations.

Automobiles provide comfort and mobility to owners. While they make life more meaningful they also pose challenges and risks in their safety and security mechanisms. Some modern automobiles are equipped with anti-theft systems and enhanced safety measures to safeguard its drivers. But at times, these mechanisms for safety and secured operation of automobiles are insufficient due to various mechanisms used by intruders and car thieves to defeat them. Drunk drivers cause accidents on our roads and thus the need to safeguard the driver when he is intoxicated and render the car to be incapable of being driven. These issues merit an integrated approach to safety and security of automobiles. In the light of these challenges, an integrated microcontroller-based hardware and software system for safety and security of automobiles to be fixed into existing vehicle architecture, was designed, developed and deployed. The system submodules are: (1) Two-step ignition for automobiles, namely: (a) biometric ignition and (b) alcohol detection with engine control, (2) Global Positioning System (GPS) based vehicle tracking and (3) Multisensor-based fire detection using neuro-fuzzy logic. All submodules of the system were implemented using one microcontroller, the Arduino Mega 2560, as the central control unit. The microcontroller was programmed using C++11. The developed system performed quite well with the tests performed on it. Given the right conditions, the alcohol detection subsystem operated with a 92% efficiency. The biometric ignition subsystem operated with about 80% efficiency. The fire detection subsystem operated with a 95% efficiency in locations registered with the neuro-fuzzy system. The vehicle tracking subsystem operated with an efficiency of 90%.

Botnet is one of the major threats on the Internet for committing cybercrimes, such as DDoS attacks, stealing sensitive information, spreading spams, etc. It is a challenging issue to detect modern botnets that are continuously improving for evading detection. In this paper, we propose a machine learning based botnet detection system that is shown to be effective in identifying P2P botnets. Our approach extracts convolutional version of effective flow-based features, and trains a classification model by using a feed-forward artificial neural network. The experimental results show that the accuracy of detection using the convolutional features is better than the ones using the traditional features. It can achieve 94.7% of detection accuracy and 2.2% of false positive rate on the known P2P botnet datasets. Furthermore, our system provides an additional confidence testing for enhancing performance of botnet detection. It further classifies the network traffic of insufficient confidence in the neural network. The experiment shows that this stage can increase the detection accuracy up to 98.6% and decrease the false positive rate up to 0.5%.

The Internet of Things (IoT) is the network where physical devices, sensors, appliances and other different objects can communicate with each other without the need for human intervention. Wireless Sensor Networks (WSNs) are main building blocks of the IoT. Both the IoT and WSNs have many critical and non-critical applications that touch almost every aspect of our modern life. Unfortunately, these networks are prone to various types of security threats. Therefore, the security of IoT and WSNs became crucial. Furthermore, the resource limitations of the devices used in these networks complicate the problem. One of the most recent and effective approaches to address such challenges is machine learning. Machine learning inspires many solutions to secure the IoT and WSNs. In this paper, we survey the different threats that can attack both IoT and WSNs and the machine learning techniques developed to counter them.

The recently developed deep belief network (DBN) has been shown to be an effective methodology for solving time series forecasting problems. However, the performance of DBN is seriously depended on the reasonable setting of hyperparameters. At present, random search, grid search and Bayesian optimization are the most common methods of hyperparameters optimization. As an alternative, a state-of-the-art derivative-free optimizer-negative correlation search (NCS) is adopted in this paper to decide the sizes of DBN and learning rates during the training processes. A comparative analysis is performed between the proposed method and other popular techniques in the time series forecasting experiment based on two types of time series datasets. Experiment results statistically affirm the efficiency of the proposed model to obtain better prediction results compared with conventional neural network models.