Matching Items (150)
Filtering by
- All Subjects: Computer Science
- Creators: Ahn, Gail-Joon
- Creators: Ye, Jieping
Description
Recently, a well-designed and well-trained neural network can yield state-of-the-art results across many domains, including data mining, computer vision, and medical image analysis. But progress has been limited for tasks where labels are difficult or impossible to obtain. This reliance on exhaustive labeling is a critical limitation in the rapid deployment of neural networks. Besides, the current research scales poorly to a large number of unseen concepts and is passively spoon-fed with data and supervision.
To overcome the above data scarcity and generalization issues, in my dissertation, I first propose two unsupervised conventional machine learning algorithms, hyperbolic stochastic coding, and multi-resemble multi-target low-rank coding, to solve the incomplete data and missing label problem. I further introduce a deep multi-domain adaptation network to leverage the power of deep learning by transferring the rich knowledge from a large-amount labeled source dataset. I also invent a novel time-sequence dynamically hierarchical network that adaptively simplifies the network to cope with the scarce data.
To learn a large number of unseen concepts, lifelong machine learning enjoys many advantages, including abstracting knowledge from prior learning and using the experience to help future learning, regardless of how much data is currently available. Incorporating this capability and making it versatile, I propose deep multi-task weight consolidation to accumulate knowledge continuously and significantly reduce data requirements in a variety of domains. Inspired by the recent breakthroughs in automatically learning suitable neural network architectures (AutoML), I develop a nonexpansive AutoML framework to train an online model without the abundance of labeled data. This work automatically expands the network to increase model capability when necessary, then compresses the model to maintain the model efficiency.
In my current ongoing work, I propose an alternative method of supervised learning that does not require direct labels. This could utilize various supervision from an image/object as a target value for supervising the target tasks without labels, and it turns out to be surprisingly effective. The proposed method only requires few-shot labeled data to train, and can self-supervised learn the information it needs and generalize to datasets not seen during training.
To overcome the above data scarcity and generalization issues, in my dissertation, I first propose two unsupervised conventional machine learning algorithms, hyperbolic stochastic coding, and multi-resemble multi-target low-rank coding, to solve the incomplete data and missing label problem. I further introduce a deep multi-domain adaptation network to leverage the power of deep learning by transferring the rich knowledge from a large-amount labeled source dataset. I also invent a novel time-sequence dynamically hierarchical network that adaptively simplifies the network to cope with the scarce data.
To learn a large number of unseen concepts, lifelong machine learning enjoys many advantages, including abstracting knowledge from prior learning and using the experience to help future learning, regardless of how much data is currently available. Incorporating this capability and making it versatile, I propose deep multi-task weight consolidation to accumulate knowledge continuously and significantly reduce data requirements in a variety of domains. Inspired by the recent breakthroughs in automatically learning suitable neural network architectures (AutoML), I develop a nonexpansive AutoML framework to train an online model without the abundance of labeled data. This work automatically expands the network to increase model capability when necessary, then compresses the model to maintain the model efficiency.
In my current ongoing work, I propose an alternative method of supervised learning that does not require direct labels. This could utilize various supervision from an image/object as a target value for supervising the target tasks without labels, and it turns out to be surprisingly effective. The proposed method only requires few-shot labeled data to train, and can self-supervised learn the information it needs and generalize to datasets not seen during training.
ContributorsZhang, Jie (Author) / Wang, Yalin (Thesis advisor) / Liu, Huan (Committee member) / Stonnington, Cynthia (Committee member) / Liang, Jianming (Committee member) / Yang, Yezhou (Committee member) / Arizona State University (Publisher)
Created2020
Description
Cyberspace has become a field where the competitive arms race between defenders and adversaries play out. Adaptive, intelligent adversaries are crafting new responses to the advanced defenses even though the arms race has resulted in a gradual improvement of the security posture. This dissertation aims to assess the evolving threat landscape and enhance state-of-the-art defenses by exploiting and mitigating two different types of emerging security vulnerabilities. I first design a new cache attack method named Prime+Count which features low noise and no shared memory needed.I use the method to construct fast data covert channels. Then, I propose a novel software-based approach, SmokeBomb, to prevent cache side-channel attacks for inclusive and non-inclusive caches based on the creation of a private space in the L1 cache. I demonstrate the effectiveness of SmokeBomb by applying it to two different ARM processors with different instruction set versions and cache models and carry out an in-depth evaluation. Next, I introduce an automated approach that exploits a stack-based information leak vulnerability in operating system kernels to obtain sensitive data. Also, I propose a lightweight and widely applicable runtime defense, ViK, for preventing temporal memory safety violations which can lead attackers to have arbitrary code execution or privilege escalation together with information leak vulnerabilities. The security impact of temporal memory safety vulnerabilities is critical, but,they are difficult to identify because of the complexity of real-world software and the spatial separation of allocation and deallocation code. Therefore, I focus on preventing not the vulnerabilities themselves, but their exploitation. ViK can effectively protect operating system kernels and user-space programs from temporal memory safety violations, imposing low runtime and memory overhead.
ContributorsCho, Haehyun (Author) / Ahn, Gail-Joon (Thesis advisor) / Doupe, Adam (Thesis advisor) / Shoshitaishvili, Yan (Committee member) / Wang, Ruoyu (Committee member) / Wu, Carole-Jean (Committee member) / Arizona State University (Publisher)
Created2021
Description
Statistical Shape Modeling is widely used to study the morphometrics of deformable objects in computer vision and biomedical studies. There are mainly two viewpoints to understand the shapes. On one hand, the outer surface of the shape can be taken as a two-dimensional embedding in space. On the other hand, the outer surface along with its enclosed internal volume can be taken as a three-dimensional embedding of interests. Most studies focus on the surface-based perspective by leveraging the intrinsic features on the tangent plane. But a two-dimensional model may fail to fully represent the realistic properties of shapes with both intrinsic and extrinsic properties. In this thesis, severalStochastic Partial Differential Equations (SPDEs) are thoroughly investigated and several methods are originated from these SPDEs to try to solve the problem of both two-dimensional and three-dimensional shape analyses. The unique physical meanings of these SPDEs inspired the findings of features, shape descriptors, metrics, and kernels in this series of works. Initially, the data generation of high-dimensional shapes, here, the tetrahedral meshes, is introduced. The cerebral cortex is taken as the study target and an automatic pipeline of generating the gray matter tetrahedral mesh is introduced. Then, a discretized Laplace-Beltrami operator (LBO) and a Hamiltonian operator (HO) in tetrahedral domain with Finite Element Method (FEM) are derived. Two high-dimensional shape descriptors are defined based on the solution of the heat equation and Schrödinger’s equation. Considering the fact that high-dimensional shape models usually contain massive redundancies, and the demands on effective landmarks in many applications, a Gaussian process landmarking on tetrahedral meshes is further studied. A SIWKS-based metric space is used to define a geometry-aware Gaussian process. The study of the periodic potential diffusion process further inspired the idea of a new kernel call the geometry-aware convolutional kernel. A series of Bayesian learning methods are then introduced to tackle the problem of shape retrieval and classification. Experiments of every single item are demonstrated. From the popular SPDE such as the heat equation and Schrödinger’s equation to the general potential diffusion equation and the specific periodic potential diffusion equation, it clearly shows that classical SPDEs play an important role in discovering new features, metrics, shape descriptors and kernels. I hope this thesis could be an example of using interdisciplinary knowledge to solve problems.
ContributorsFan, Yonghui (Author) / Wang, Yalin (Thesis advisor) / Lepore, Natasha (Committee member) / Turaga, Pavan (Committee member) / Yang, Yezhou (Committee member) / Arizona State University (Publisher)
Created2021
Description
Emerging trends in cyber system security breaches in critical cloud infrastructures show that attackers have abundant resources (human and computing power), expertise and support of large organizations and possible foreign governments. In order to greatly improve the protection of critical cloud infrastructures, incorporation of human behavior is needed to predict potential security breaches in critical cloud infrastructures. To achieve such prediction, it is envisioned to develop a probabilistic modeling approach with the capability of accurately capturing system-wide causal relationship among the observed operational behaviors in the critical cloud infrastructure and accurately capturing probabilistic human (users’) behaviors on subsystems as the subsystems are directly interacting with humans. In our conceptual approach, the system-wide causal relationship can be captured by the Bayesian network, and the probabilistic human behavior in the subsystems can be captured by the Markov Decision Processes. The interactions between the dynamically changing state graphs of Markov Decision Processes and the dynamic causal relationships in Bayesian network are key components in such probabilistic modelling applications. In this thesis, two techniques are presented for supporting the above vision to prediction of potential security breaches in critical cloud infrastructures. The first technique is for evaluation of the conformance of the Bayesian network with the multiple MDPs. The second technique is to evaluate the dynamically changing Bayesian network structure for conformance with the rules of the Bayesian network using a graph checker algorithm. A case study and its simulation are presented to show how the two techniques support the specific parts in our conceptual approach to predicting system-wide security breaches in critical cloud infrastructures.
ContributorsNagaraja, Vinjith (Author) / Yau, Stephen S. (Thesis advisor) / Ahn, Gail-Joon (Committee member) / Davulcu, Hasan (Committee member) / Arizona State University (Publisher)
Created2015
Description
One of the most remarkable outcomes resulting from the evolution of the web into Web 2.0, has been the propelling of blogging into a widely adopted and globally accepted phenomenon. While the unprecedented growth of the Blogosphere has added diversity and enriched the media, it has also added complexity. To cope with the relentless expansion, many enthusiastic bloggers have embarked on voluntarily writing, tagging, labeling, and cataloguing their posts in hopes of reaching the widest possible audience. Unbeknown to them, this reaching-for-others process triggers the generation of a new kind of collective wisdom, a result of shared collaboration, and the exchange of ideas, purpose, and objectives, through the formation of associations, links, and relations. Mastering an understanding of the Blogosphere can greatly help facilitate the needs of the ever growing number of these users, as well as producers, service providers, and advertisers into facilitation of the categorization and navigation of this vast environment. This work explores a novel method to leverage the collective wisdom from the infused label space for blog search and discovery. The work demonstrates that the wisdom space can provide a most unique and desirable framework to which to discover the highly sought after background information that could aid in the building of classifiers. This work incorporates this insight into the construction of a better clustering of blogs which boosts the performance of classifiers for identifying more relevant labels for blogs, and offers a mechanism that can be incorporated into replacing spurious labels and mislabels in a multi-labeled space.
ContributorsGalan, Magdiel F (Author) / Liu, Huan (Thesis advisor) / Davulcu, Hasan (Committee member) / Ye, Jieping (Committee member) / Li, Baoxin (Committee member) / Arizona State University (Publisher)
Created2015
Description
Cyber threats are growing in number and sophistication making it important to continually study and improve all dimensions of cyber defense. Human teamwork in cyber defense analysis has been overlooked even though it has been identified as an important predictor of cyber defense performance. Also, to detect advanced forms of threats effective information sharing and collaboration between the cyber defense analysts becomes imperative. Therefore, through this dissertation work, I took a cognitive engineering approach to investigate and improve cyber defense teamwork. The approach involved investigating a plausible team-level bias called the information pooling bias in cyber defense analyst teams conducting the detection task that is part of forensics analysis through human-in-the-loop experimentation. The approach also involved developing agent-based models based on the experimental results to explore the cognitive underpinnings of this bias in human analysts. A prototype collaborative visualization tool was developed by considering the plausible cognitive limitations contributing to the bias to investigate whether a cognitive engineering-driven visualization tool can help mitigate the bias in comparison to off-the-shelf tools. It was found that participant teams conducting the collaborative detection tasks as part of forensics analysis, experience the information pooling bias affecting their performance. Results indicate that cognitive friendly visualizations can help mitigate the effect of this bias in cyber defense analysts. Agent-based modeling produced insights on internal cognitive processes that might be contributing to this bias which could be leveraged in building future visualizations. This work has multiple implications including the development of new knowledge about the science of cyber defense teamwork, a demonstration of the advantage of developing tools using a cognitive engineering approach, a demonstration of the advantage of using a hybrid cognitive engineering methodology to study teams in general and finally, a demonstration of the effect of effective teamwork on cyber defense performance.
ContributorsRajivan, Prashanth (Author) / Cooke, Nancy J. (Thesis advisor) / Ahn, Gail-Joon (Committee member) / Janssen, Marcus (Committee member) / Arizona State University (Publisher)
Created2014
Description
Mobile Augmented Reality (MAR) is a portable, powerful, and suitable technology that integrates 3D virtual content into the physical world in real-time. It has been implemented for multiple intents as it enhances people’s interaction, e.g., shopping, entertainment, gaming, etc. Thus, MAR is expected to grow at a tremendous rate in the upcoming years, as its popularity via mobile devices has increased. But, unfortunately, the applications that implement MAR, hereby referred to as MAR-Apps, bear security issues. Such are imaged in worldwide recorded incidents caused by MAR-Apps, e.g., robberies, authorities requesting banning MAR at specific locations, etc. To further explore these concerns, a case study analyzed several MAR-Apps available in the market to identify the security problems in MAR. As a result of this study, the threats found were classified into three categories. First, Space Invasion implies the intrusive modification through MAR of sensitive spaces, e.g., hospitals, memorials, etc. Then, Space Affectation means the degradation of users’ experience via interaction with undesirable MAR or malicious entities. Finally, MAR-Apps mishandling sensitive data leads to Privacy Leaks. SpaceMediator, a proof-of-concept MAR-App that imitates the well-known and successful MAR-App Pokémon GO, implements the solution approach of a Policy-Governed MAR-App, which assists in preventing the aforementioned mentioned security issues. Furthermore, its feasibility is evaluated through a user study with 40 participants. As a result, uncovering understandability over the security issues as participants recognized and prevented them with success rates as high as 92.50%. Furthermore, there is an enriched interest in Policy-Governed MAR-Apps as 87.50% of participants agreed with restricted MAR-Apps within sensitive spaces, and 82.50% would implement constraints in MAR-Apps. These promising results encourage adopting the Policy-Governed solution approach in future MAR-Apps.
ContributorsClaramunt, Luis Manuel (Author) / Ahn, Gail-Joon (Thesis advisor) / Rubio-Medrano, Carlos E (Committee member) / Baek, Jaejong (Committee member) / Arizona State University (Publisher)
Created2022
Description
Studies on underground forums can significantly advance the understanding of cybercrime workflow and underground economies. However, research on underground forums has concentrated on public information with little attention paid to users’ private interactions. Since detailed information will be discussed privately, the failure to investigate private interactions may miss critical intelligence and even misunderstand the entire underground economy. Furthermore, underground forums have evolved into criminal freelance markets where criminals trade illicit products and cybercrime services, allowing unsophisticated people to launch sophisticated cyber attacks. However, current research rarely examines and explores how criminals interact with each other, which makes researchers miss the opportunities to detect new cybercrime patterns proactively. Moreover, in clearnet, criminals are active in exploiting human vulnerabilities to conduct various attacks, and the phishing attack is one of the most prevalent types of cybercrime. Phishing awareness training has been proven to decrease the rate of clicking phishing emails. However, the rate of reporting phishing attacks is unexpectedly low based on recent studies, leaving phishing websites with hours of additional active time before being detected. In this dissertation, I first present an analysis of private interactions in underground forums and introduce machine learning-based approaches to detect hidden connections between users. Secondly, I analyze how criminals collaborate with each other in an emerging scam service in underground forums that exploits the return policies of merchants to get a refund or a replacement without returning the purchased products. Finally, I conduct a comprehensive evaluation of the phishing reporting ecosystem to identify the critical challenges while reporting phishing attacks to enable people to fight against phishers proactively.
ContributorsSun, Zhibo (Author) / Ahn, Gail-Joon (Thesis advisor) / Doupe, Adam (Thesis advisor) / Bao, Tiffany (Committee member) / Benjamin, Victor (Committee member) / Arizona State University (Publisher)
Created2022
Description
Drosophila melanogaster, as an important model organism, is used to explore the mechanism which governs cell differentiation and embryonic development. Understanding the mechanism will help to reveal the effects of genes on other species or even human beings. Currently, digital camera techniques make high quality Drosophila gene expression imaging possible. On the other hand, due to the advances in biology, gene expression images which can reveal spatiotemporal patterns are generated in a high-throughput pace. Thus, an automated and efficient system that can analyze gene expression will become a necessary tool for investigating the gene functions, interactions and developmental processes. One investigation method is to compare the expression patterns of different developmental stages. Recently, however, the expression patterns are manually annotated with rough stage ranges. The work of annotation requires professional knowledge from experienced biologists. Hence, how to transfer the domain knowledge in biology into an automated system which can automatically annotate the patterns provides a challenging problem for computer scientists. In this thesis, the problem of stage annotation for Drosophila embryo is modeled in the machine learning framework. Three sparse learning algorithms and one ensemble algorithm are used to attack the problem. The sparse algorithms are Lasso, group Lasso and sparse group Lasso. The ensemble algorithm is based on a voting method. Besides that the proposed algorithms can annotate the patterns to stages instead of stage ranges with high accuracy; the decimal stage annotation algorithm presents a novel way to annotate the patterns to decimal stages. In addition, some analysis on the algorithm performance are made and corresponding explanations are given. Finally, with the proposed system, all the lateral view BDGP and FlyFish images are annotated and several interesting applications of decimal stage value are revealed.
ContributorsPan, Cheng (Author) / Ye, Jieping (Thesis advisor) / Li, Baoxin (Committee member) / Farin, Gerald (Committee member) / Arizona State University (Publisher)
Created2012
Description
Nowadays, wireless communications and networks have been widely used in our daily lives. One of the most important topics related to networking research is using optimization tools to improve the utilization of network resources. In this dissertation, we concentrate on optimization for resource-constrained wireless networks, and study two fundamental resource-allocation problems: 1) distributed routing optimization and 2) anypath routing optimization. The study on the distributed routing optimization problem is composed of two main thrusts, targeted at understanding distributed routing and resource optimization for multihop wireless networks. The first thrust is dedicated to understanding the impact of full-duplex transmission on wireless network resource optimization. We propose two provably good distributed algorithms to optimize the resources in a full-duplex wireless network. We prove their optimality and also provide network status analysis using dual space information. The second thrust is dedicated to understanding the influence of network entity load constraints on network resource allocation and routing computation. We propose a provably good distributed algorithm to allocate wireless resources. In addition, we propose a new subgradient optimization framework, which can provide findgrained convergence, optimality, and dual space information at each iteration. This framework can provide a useful theoretical foundation for many networking optimization problems. The study on the anypath routing optimization problem is composed of two main thrusts. The first thrust is dedicated to understanding the computational complexity of multi-constrained anypath routing and designing approximate solutions. We prove that this problem is NP-hard when the number of constraints is larger than one. We present two polynomial time K-approximation algorithms. One is a centralized algorithm while the other one is a distributed algorithm. For the second thrust, we study directional anypath routing and present a cross-layer design of MAC and routing. For the MAC layer, we present a directional anycast MAC. For the routing layer, we propose two polynomial time routing algorithms to compute directional anypaths based on two antenna models, and prove their ptimality based on the packet delivery ratio metric.
ContributorsFang, Xi (Author) / Xue, Guoliang (Thesis advisor) / Yau, Sik-Sang (Committee member) / Ye, Jieping (Committee member) / Zhang, Junshan (Committee member) / Arizona State University (Publisher)
Created2013