Biblio

Deep neural networks (DNNs) provide good performance for image recognition, speech recognition, and pattern recognition. However, a poisoning attack is a serious threat to DNN's security. The poisoning attack is a method to reduce the accuracy of DNN by adding malicious training data during DNN training process. In some situations such as a military, it may be necessary to drop only a chosen class of accuracy in the model. For example, if an attacker does not allow only nuclear facilities to be selectively recognized, it may be necessary to intentionally prevent UAV from correctly recognizing nuclear-related facilities. In this paper, we propose a selective poisoning attack that reduces the accuracy of only chosen class in the model. The proposed method reduces the accuracy of a chosen class in the model by training malicious training data corresponding to a chosen class, while maintaining the accuracy of the remaining classes. For experiment, we used tensorflow as a machine learning library and MNIST and CIFAR10 as datasets. Experimental results show that the proposed method can reduce the accuracy of the chosen class to 43.2% and 55.3% in MNIST and CIFAR10, while maintaining the accuracy of the remaining classes.

In this paper, we present a semi-supervised remote sensing change detection method based on graph model with Generative Adversarial Networks (GANs). Firstly, the multi-temporal remote sensing change detection problem is converted as a problem of semi-supervised learning on graph where a majority of unlabeled nodes and a few labeled nodes are contained. Then, GANs are adopted to generate samples in a competitive manner and help improve the classification accuracy. Finally, a binary change map is produced by classifying the unlabeled nodes to a certain class with the help of both the labeled nodes and the unlabeled nodes on graph. Experimental results carried on several very high resolution remote sensing image data sets demonstrate the effectiveness of our method.

E-mail is widespread and an essential communication technology in modern times. Since e-mail has problems with spam mails and spoofed e-mails, countermeasures are required. Although SPF, DKIM and DMARC have been proposed as sender domain authentication, these mechanisms cannot detect non-spoofing spam mails. To overcome this issue, this paper proposes a method to detect spam domains by supervised learning with features extracted from e-mail reception log and active DNS data, such as the result of Sender Authentication, the Sender IP address, the number of each DNS record, and so on. As a result of the experiment, our method can detect spam domains with 88.09% accuracy and 97.11% precision. We confirmed that our method can detect spam domains with detection accuracy 19.40% higher than the previous study by utilizing not only active DNS data but also e-mail reception log in combination.

Static vulnerability detection has shown its effectiveness in detecting well-defined low-level memory errors. However, high-level control-flow related (CFR) vulnerabilities, such as insufficient control flow management (CWE-691), business logic errors (CWE-840), and program behavioral problems (CWE-438), which are often caused by a wide variety of bad programming practices, posing a great challenge for existing general static analysis solutions. This paper presents a new deep-learning-based graph embedding approach to accurate detection of CFR vulnerabilities. Our approach makes a new attempt by applying a recent graph convolutional network to embed code fragments in a compact and low-dimensional representation that preserves high-level control-flow information of a vulnerable program. We have conducted our experiments using 8,368 real-world vulnerable programs by comparing our approach with several traditional static vulnerability detectors and state-of-the-art machine-learning-based approaches. The experimental results show the effectiveness of our approach in terms of both accuracy and recall. Our research has shed light on the promising direction of combining program analysis with deep learning techniques to address the general static analysis challenges.

Internet application providers now have more incentive than ever to collect user data, which greatly increases the risk of user privacy violations due to the emerging of deep neural networks. In this paper, we propose TensorClog-a poisoning attack technique that is designed for privacy protection against deep neural networks. TensorClog has three properties with each of them serving a privacy protection purpose: 1) training on TensorClog poisoned data results in lower inference accuracy, reducing the incentive of abusive data collection; 2) training on TensorClog poisoned data converges to a larger loss, which prevents the neural network from learning the privacy; and 3) TensorClog regularizes the perturbation to remain a high structure similarity, so that the poisoning does not affect the actual content in the data. Applying our TensorClog poisoning technique to CIFAR-10 dataset results in an increase in both converged training loss and test error by 300% and 272%, respectively. It manages to maintain data's human perception with a high SSIM index of 0.9905. More experiments including different limited information attack scenarios and a real-world application transferred from pre-trained ImageNet models are presented to further evaluate TensorClog's effectiveness in more complex situations.

This study has built a simulation of a smart home system by the Alibaba ECS. The architecture of hardware was based on edge computing technology. The whole method would design a clear classifier to find the boundary between regular and mutation codes. It could be applied in the detection of the mutation code of network. The project has used the dataset vector to divide them into positive and negative type, and the final result has shown the RBF-function SVM method perform best in this mission. This research has got a good network security detection in the IoT systems and increased the applications of machine learning.

The goal of content-based recommendation system is to retrieve and rank the list of items that are closest to the query item. Today, almost every e-commerce platform has a recommendation system strategy for products that customers can decide to buy. In this paper we describe our work on creating a Generative Adversarial Network based image retrieval system for e-commerce platforms to retrieve best similar images for a given product image specifically for shoes. We compare state-of-the-art solutions and provide results for the proposed deep learning network on a standard data set.

In this paper, we explore the use of machine learning technique for wormhole attack detection in ad hoc network. This work has categorized into three major tasks. One of our tasks is a simulation of wormhole attack in an ad hoc network environment with multiple wormhole tunnels. A next task is the characterization of packet attributes that lead to feature selection. Consequently, we perform data generation and data collection operation that provide large volume dataset. The final task is applied to machine learning technique for wormhole attack detection. Prior to this, a wormhole attack has detected using traditional approaches. In those, a Multirate-DelPHI is shown best results as detection rate is 90%, and the false alarm rate is 20%. We conduct experiments and illustrate that our method performs better resulting in all statistical parameters such as detection rate is 93.12% and false alarm rate is 5.3%. Furthermore, we have also shown results on various statistical parameters such as Precision, F-measure, MCC, and Accuracy.

Style transfer is a research hotspot in computer vision. Up to now, it is still a challenge although many researches have been conducted on it for high quality style transfer. In this work, we propose an algorithm named ASTCNN which is a real-time Arbitrary Style Transfer Convolution Neural Network. The ASTCNN consists of two independent encoders and a decoder. The encoders respectively extract style and content features from style and content and the decoder generates the style transferred image images. Experimental results show that ASTCNN achieves higher quality output image than the state-of-the-art style transfer algorithms and the floating point computation of ASTCNN is 23.3% less than theirs.

In order to solve the problem that there is no effective means to find the optimal number of hidden nodes of single-hidden-layer feedforward neural network, in this paper, a method will be introduced to solve it effectively by using singular value decomposition. First, the training data need to be normalized strictly by attribute-based data normalization and sample-based data normalization. Then, the normalized data is decomposed based on the singular value decomposition, and the number of hidden nodes is determined according to main eigenvalues. The experimental results of MNIST data set and APS data set show that the feedforward neural network can attain satisfactory performance in the classification task.

With the tremendous growth of IoT botnet DDoS attacks in recent years, IoT security has now become one of the most concerned topics in the field of network security. A lot of security approaches have been proposed in the area, but they still lack in terms of dealing with newer emerging variants of IoT malware, known as Zero-Day Attacks. In this paper, we present a honeypot-based approach which uses machine learning techniques for malware detection. The IoT honeypot generated data is used as a dataset for the effective and dynamic training of a machine learning model. The approach can be taken as a productive outset towards combatting Zero-Day DDoS Attacks which now has emerged as an open challenge in defending IoT against DDoS Attacks.

The continuing decrease in feature size of integrated circuits, and the increase of the complexity and cost of design and fabrication has led to outsourcing the design and fabrication of integrated circuits to third parties across the globe, and in turn has introduced several security vulnerabilities. The adversaries in the supply chain can pirate integrated circuits, overproduce these circuits, perform reverse engineering, and/or insert hardware Trojans in these circuits. Developing countermeasures against such security threats is highly crucial. Accordingly, this paper first develops a learning-based trust verification framework to detect hardware Trojans. To tackle Trojan insertion, IP piracy and overproduction, logic locking schemes and in particular stripped functionality logic locking is discussed and its resiliency against the state-of-the-art attacks is investigated.

In recent years, sentiment analysis has been aware within software engineering domain. While automated sentiment analysis has long been suffering from doubt of accuracy, the tool performance is unstable when being applied on datasets other than the original dataset for evaluation. Researchers also have the disagreements upon if machine learning algorithms perform better than conventional lexicon and rule based approaches. In this paper, we looked into the factors in datasets that may affect the evaluation performance, also evaluated the popular machine learning algorithms in sentiment analysis, then proposed a novel structure for automated sentiment tool combines advantages from both approaches.

To solve the high-resolution three-dimensional (3D) microwave imaging is a challenging topic due to its inherent unmanageable computation. Recently, deep learning techniques that can fully explore the prior of meaningful pattern embodied in data have begun to show its intriguing merits in various areas of inverse problem. Motivated by this observation, we here present a deep-learning-inspired approach to the high-resolution 3D microwave imaging in the context of Generative Adversarial Network (GAN), termed as GANMI in this work. Simulation and experimental results have been provided to demonstrate that the proposed GANMI can remarkably outperform conventional methods in terms of both the image quality and computational time.

Fingerprinting the malware by its behavioural signature has been an attractive approach for malware detection due to the homogeneity of dynamic execution patterns across different variants of similar families. Although previous researches show reasonably good performance in dynamic detection using machine learning techniques on a large corpus of training set, decisions must be undertaken based upon a scarce number of observable samples in many practical defence scenarios. This paper demonstrates the effectiveness of generative adversarial autoencoder for dynamic malware detection under outbreak situations where in most cases a single sample is available for training the machine learning algorithm to detect similar samples that are in the wild.

Neural Network is used in many applications and guarding its performance against faults is a research challenge. Self-healing neural network is a promising concept for achieving reliability, which is the ability to detect and fix a fault in the system automatically. Most of the current self-healing neural network are based on replication of hardware nodes which causes significant area overhead. The proposed self-healing approach results in a modest area overhead and it is suitable for complex neural network. The proposed method is based on a shared operation and a spare node in each layer which compensates for any faulty node in the layer. Each faulty node will be compensated by its neighbor node, and the neighbor node performs the faulty node as well as its own operations sequentially. In the case the neighbor is faulty, the spare node will compensate for it. The proposed method is implemented using VHDL and the simulation results are obtained using Altira 10 GX FPGA for a different number of nodes. The area overhead is very small for a complex network. The reliability of the proposed method is studied and compared with the traditional neural network.

Federated learning is a novel distributed learning framework, where the deep learning model is trained in a collaborative manner among thousands of participants. The shares between server and participants are only model parameters, which prevent the server from direct access to the private training data. However, we notice that the federated learning architecture is vulnerable to an active attack from insider participants, called poisoning attack, where the attacker can act as a benign participant in federated learning to upload the poisoned update to the server so that he can easily affect the performance of the global model. In this work, we study and evaluate a poisoning attack in federated learning system based on generative adversarial nets (GAN). That is, an attacker first acts as a benign participant and stealthily trains a GAN to mimic prototypical samples of the other participants' training set which does not belong to the attacker. Then these generated samples will be fully controlled by the attacker to generate the poisoning updates, and the global model will be compromised by the attacker with uploading the scaled poisoning updates to the server. In our evaluation, we show that the attacker in our construction can successfully generate samples of other benign participants using GAN and the global model performs more than 80% accuracy on both poisoning tasks and main tasks.

These days deep learning is the fastest-growing area in the field of Machine Learning. Convolutional Neural Networks are currently the main tool used for the image analysis and classification purposes. Although great achievements and perspectives, deep neural networks and accompanying learning algorithms have some relevant challenges to tackle. In this paper, we have focused on the most frequently mentioned problem in the field of machine learning, that is relatively poor generalization abilities. Partial remedies for this are regularization techniques e.g. dropout, batch normalization, weight decay, transfer learning, early stopping and data augmentation. In this paper we have focused on data augmentation. We propose to use a method based on a neural style transfer, which allows to generate new unlabeled images of high perceptual quality that combine the content of a base image with the appearance of another one. In a proposed approach, the newly created images are described with pseudo-labels, and then used as a training dataset. Real, labeled images are divided into the validation and test set. We validated proposed method on a challenging skin lesion classification case study. Four representative neural architectures are examined. Obtained results show the strong potential of the proposed approach.

This paper introduces a biologically inspired associative data model and structure for finding nearest neighbors and similar patterns. The method can be used as an alternative to the classical approaches to accelerate the search for such patterns using various priorities for attributes according to the Sebestyen measure. The presented structure, together with algorithms developed in this paper can be useful in various computational intelligence tasks like pattern matching, recognition, clustering, classification, multi-criterion search etc. This approach is particularly useful for the on-line operation of associative neural network graphs. Graphs that dynamically develop their structure during learning on training data. The results of experiments show that the associative approach can substantially accelerate the nearest neighbor search and that associative structures can also be used as a model for KNN tasks. Finally, this paper presents how the associative structures can be used to self-organize data and represent knowledge about them in the associative way, which yields new search approaches described in this paper.

The concept of a decoy process is a new development of defensive deception beyond traditional honeypots. Decoy processes can be exceptionally effective in detecting malware, directly upon contact or by redirecting malware to decoy I/O. A key requirement is that they resemble their real counterparts very closely to withstand adversarial probes by threat actors. To be usable, decoy processes need to consume only a small fraction of the resources consumed by their real counterparts. Our contribution in this paper is twofold. We attack the resource utilization consistency of decoy processes provided by a neural network with a heatmap training mechanism, which we find to be insufficiently trained. We then devise machine learning over control flow graphs that improves the heatmap training mechanism. A neural network retrained by our work shows higher accuracy and defeats our attacks without a significant increase in its own resource utilization.

Deep learning is a popular powerful machine learning solution to the computer vision tasks. The most criticized vulnerability of deep learning is its poor tolerance towards adversarial images obtained by deliberately adding imperceptibly small perturbations to the clean inputs. Such negatives can delude a classifier into wrong decision making. Previous defensive techniques mostly focused on refining the models or input transformation. They are either implemented only with small datasets or shown to have limited success. Furthermore, they are rarely scrutinized from the hardware perspective despite Artificial Intelligence (AI) on a chip is a roadmap for embedded intelligence everywhere. In this paper we propose a new discriminative noise injection strategy to adaptively select a few dominant layers and progressively discriminate adversarial from benign inputs. This is made possible by evaluating the differences in label change rate from both adversarial and natural images by injecting different amount of noise into the weights of individual layers in the model. The approach is evaluated on the ImageNet Dataset with 8-bit truncated models for the state-of-the-art DNN architectures. The results show a high detection rate of up to 88.00% with only approximately 5% of false positive rate for MobileNet. Both detection rate and false positive rate have been improved well above existing advanced defenses against the most practical noninvasive universal perturbation attack on deep learning based AI chip.

Adversarial attacks can affect the performance of existing deep learning models. With the increased interest in graph based machine learning techniques, there have been investigations which suggest that these models are also vulnerable to attacks. In particular, corruptions of the graph topology can degrade the performance of graph based learning algorithms severely. This is due to the fact that the prediction capability of these algorithms relies mostly on the similarity structure imposed by the graph connectivity. Therefore, detecting the location of the corruption and correcting the induced errors becomes crucial. There has been some recent work which tackles the detection problem, however these methods do not address the effect of the attack on the downstream learning task. In this work, we propose an algorithm that uses node copying to mitigate the degradation in classification that is caused by adversarial attacks. The proposed methodology is applied only after the model for the downstream task is trained and the added computation cost scales well for large graphs. Experimental results show the effectiveness of our approach for several real world datasets.

Twitter being the most popular content sharing platform is giving rise to automated accounts called “bots”. Majority of the users on Twitter are bots. Various machine learning (ML) algorithms are designed to detect bots avoiding the vulnerability constraints of ML-based models. This paper contributes to exploit vulnerabilities of machine learning (ML) algorithms through black-box attack. An adversarial text sequence misclassifies the results of deep learning (DL) classifiers for bot detection. Literature shows that ML models are vulnerable to attacks. The aim of this paper is to compromise the accuracy of ML-based bot detection algorithms by replacing original words in tweets with their synonyms. Our results show 7.2% decrease in the accuracy for bot tweets, therefore classifying bot tweets as legitimate tweets.

Inference of unknown opinions with uncertain, adversarial (e.g., incorrect or conflicting) evidence in large datasets is not a trivial task. Without proper handling, it can easily mislead decision making in data mining tasks. In this work, we propose a highly scalable opinion inference probabilistic model, namely Adversarial Collective Opinion Inference (Adv-COI), which provides a solution to infer unknown opinions with high scalability and robustness under the presence of uncertain, adversarial evidence by enhancing Collective Subjective Logic (CSL) which is developed by combining SL and Probabilistic Soft Logic (PSL). The key idea behind the Adv-COI is to learn a model of robust ways against uncertain, adversarial evidence which is formulated as a min-max problem. We validate the out-performance of the Adv-COI compared to baseline models and its competitive counterparts under possible adversarial attacks on the logic-rule based structured data and white and black box adversarial attacks under both clean and perturbed semi-synthetic and real-world datasets in three real world applications. The results show that the Adv-COI generates the lowest mean absolute error in the expected truth probability while producing the lowest running time among all.

Network attack is a significant security issue for modern society. From small mobile devices to large cloud platforms, almost all computing products, used in our daily life, are networked and potentially under the threat of network intrusion. With the fast-growing network users, network intrusions become more and more frequent, volatile and advanced. Being able to capture intrusions in time for such a large scale network is critical and very challenging. To this end, the machine learning (or AI) based network intrusion detection (NID), due to its intelligent capability, has drawn increasing attention in recent years. Compared to the traditional signature-based approaches, the AI-based solutions are more capable of detecting variants of advanced network attacks. However, the high detection rate achieved by the existing designs is usually accompanied by a high rate of false alarms, which may significantly discount the overall effectiveness of the intrusion detection system. In this paper, we consider the existence of spatial and temporal features in the network traffic data and propose a hierarchical CNN+RNN neural network, LuNet. In LuNet, the convolutional neural network (CNN) and the recurrent neural network (RNN) learn input traffic data in sync with a gradually increasing granularity such that both spatial and temporal features of the data can be effectively extracted. Our experiments on two network traffic datasets show that compared to the state-of-the-art network intrusion detection techniques, LuNet not only offers a high level of detection capability but also has a much low rate of false positive-alarm.