Digital Commons Network^™

Face Recognition Using Principal Component Analysis, Timothy Kevin Larkin

Theses

Current methods of face recognition use linear methods to extract features. This causes potentially valuable nonlinear features to be lost. Using a kernel to extract nonlinear features should lead to better feature extraction and, therefore, lower error rates. Kernel Principal Component Analysis (KPCA) will be used as the method for nonlinear feature extraction. KPCA will be compared with well known linear methods such as correlation, Eigenfaces, and Fisherfaces.

An Experimental Study On Network Intrusion Detection Systems, Peng Fu

Theses

A signature database is the key component of an elaborate intrusion detection system. The efficiency of signature generation for an intrusion detection system is a crucial requirement because of the rapid appearance of new attacks on the World Wide Web. However, in the commercial applications, signature generation is still a manual process, which requires professional skills and heavy human effort. Knowledge Discovery and Data Mining methods may be a solution to this problem. Data Mining and Machine Learning algorithms can be applied to the network traffic databases, in order to automatically generate signatures.

The purpose of this thesis and the …

Bear: An Open-Source Virtual Secure Coprocessor Based On Tcpa, Rich Macdonald, Sean Smith, John Marchesini, Omen Wild

Computer Science Technical Reports

This paper reports on our ongoing project to use TCPA to transform a desktop Linux machine into a virtual secure coprocessor: more powerful but less secure than higher-end devices. We use TCPA hardware and modified boot loaders to protect fairly static components, such as a trusted kernel; we use an enforcer module---configured as Linux Security Module---to protected more dynamic system components; we use an encrypted loopback filesystem to protect highly dynamic components. All our code is open source and available under GPL from http://enforcer.sourceforge.net/

Secure Remote Control And Configuration Of Fpx Platform In Gigabit Ethernet Environment, Haoyu Song

All Computer Science and Engineering Research

Because of its flexibility and high performance, reconfigurable logic functions implemented on the Field-programmable Port Extender (FPX ) are well suited for implementing network processing such as packet classification, filtering and intrusion detection functions. This project focuses on two key aspects of the FPX system. One is providing a Gigabit Ethernet interface by designing logic for a FPGA which is located on a line card. Address Resolution Protocol (ARP) packets are handled in hardware and Ethernet frames are processed and transformed into cells suitable for standard FPX application. The other effort is to provide a secure channel to enable remote …

Formal Properties Of Linear Memory Types, Heng Huang, Lea Wittie, Chris Hawblitzel

Computer Science Technical Reports

Efficient low-level systems need more control over memory than safe high-level languages usually provide. As a result, run-time systems are typically written in unsafe languages such as C. This report describes an abstract machine designed to give type-safe code more control over memory. It includes complete definitions and proofs of preservation, progress, strong normalization, erasure, and translation correctness.

802.11 Denial-Of-Service Attacks: Real Vulnerabilities And Practical Solutions, John M. Bellardo, Stefan Savage

Computer Science and Software Engineering

The convenience of 802.11-based wireless access networks has led to widespread deployment in the consumer, industrial and military sectors. However, this use is predicated on an implicit assumption of confidentiality and availability. While the security flaws in 802.11's basic confidentially mechanisms have been widely publicized, the threats to network availability are far less widely appreciated. In fact, it has been suggested that 802.11 is highly susceptible to malicious denial-of-service (DoS) attacks targeting its management and media access protocols. This paper provides an experimental analysis of such 802.11-specific attacks - their practicality, their efficacy and potential low-overhead implementation changes to mitigate …

Boltzmann Machine Learning With The Latent Maximum Entropy Principle, Shaojun Wang, Dale Schuurmans, Fuchun Peng, Yunxin Zhao

Kno.e.sis Publications

We present a new statistical learning paradigm for Boltzmann machines based on a new inference principle we have proposed: the latent maximum entropy principle (LME). LME is different both from Jaynes maximum entropy principle and from standard maximum likelihood estimation. We demonstrate the LME principle BY deriving new algorithms for Boltzmann machine parameter estimation, and show how robust and fast new variant of the EM algorithm can be developed. Our experiments show that estimation based on LME generally yields better results than maximum likelihood estimation, particularly when inferring hidden units from small amounts of data.

Learning Mixture Models With The Latent Maximum Entropy Principle, Shaojun Wang, Dale Schuurmans, Fuchun Peng, Yunxin Zhao

Kno.e.sis Publications

We present a new approach to estimating mixture models based on a new inference principle we have proposed: the latent maximum entropy principle (LME). LME is different both from Jaynes’ maximum entropy principle and from standard maximum likelihood estimation. We demonstrate the LME principle by deriving new algorithms for mixture model estimation, and show how robust new variants of the EM algorithm can be developed. Our experiments show that estimation based on LME generally yields better results than maximum likelihood estimation, particularly when inferring latent variable models from small amounts of data.

Enemy At The Gate: Threats To Information Security, Michael E. Whitman

Faculty Articles

A firm can build more effective security strategies by identifying and ranking the severity of potential threats to its IS efforts.

Query Protocols For Highly Resilient Peer-To-Peer Networks, Suresh Jagannathan, Gopal Pandurangan, Sriram Srinivasan

Department of Computer Science Technical Reports

No abstract provided.

Application Adaptive Bandwidth Management Using Real-Time Network Monitoring., Amit Grover

Electronic Theses and Dissertations

Application adaptive bandwidth management is a strategy for ensuring secure and reliable network operation in the presence of undesirable applications competing for a network’s crucial bandwidth, covert channels of communication via non-standard traffic on well-known ports, and coordinated Denial of Service attacks. The study undertaken here explored the classification, analysis and management of the network traffic on the basis of ports and protocols used, type of applications, traffic direction and flow rates on the East Tennessee State University’s campus-wide network. Bandwidth measurements over a nine-month period indicated bandwidth abuse of less than 0.0001% of total network bandwidth. The conclusion suggests …

Two-Dimensional Cubic Convolution, Stephen E. Reichenbach, Frank Geng

School of Computing: Faculty Publications

This paper develops two-dimensional (2-D), nonseparable, piecewise cubic convolution (PCC) for image interpolation. Traditionally, PCC has been implemented based on a one-dimensional (1-D) derivation with a separable generalization to two dimensions. However, typical scenes and imaging systems are not separable, so the traditional approach is suboptimal. We develop a closed-form derivation for a two-parameter, 2-D PCC kernel with support [-2, 2] [-2, 2] that is constrained for continuity, smoothness, symmetry, and flat-field response. Our analyses using several image models, including Markov random fields, demonstrate that the 2-D PCC yields small improvements in interpolation fidelity over the traditional, separable approach. The …

Information Systems Requirements Determination And Analysis: A Mental Modeling Approach, Keng Siau, X. Tan

Research Collection School Of Computing and Information Systems

Regardless of the system development methodology adopted, being able to understand users’ informationrequirements is vital to the success of an information systems project. The intrinsic communications obstaclesexist within an individual user, among users, and between users and systems analysts. Different requirementsdetermination and analysis (RDA) techniques have been widely used with different outcomes. Some limitations of these techniques can be identified. In this paper, we propose a mental modeling approach to support RDA. Popular in other disciplines, mental modeling techniques have some distinct advantages to overcome some intrinsic communication obstacles in RDA, and to complement other techniques.

The Effect Of Domain Knowledge On Icon Visualization, Keng Siau, Fiona Fui-Hoon Nah

Research Collection School Of Computing and Information Systems

Iconic interfaces are now the de facto interface for most computer systems. Despite the popularity of iconic interfaces and the widespread belief that iconic interfaces are easier to comprehend than non-iconic interfaces, we have not come across any published study that has examined the effect of domain knowledge on end users’ interpretation of icons. An understanding of the relationship between domain knowledge and its effect on interpretation will enable us to design better icons to not only facilitate end users’ interpretation but also reduce misinterpretations of icons. This paper reports on an experimental study that investigates the effect of domain …

The Mistaken Axioms Of Wireless-Network Research, David Kotz, Calvin Newport, Chip Elliott

Computer Science Technical Reports

Most research on ad-hoc wireless networks makes simplifying assumptions about radio propagation. The ``Flat Earth'' model of the world is surprisingly popular: all radios have circular range, have perfect coverage in that range, and travel on a two-dimensional plane. CMU's ns-2 radio models are better but still fail to represent many aspects of realistic radio networks, including hills, obstacles, link asymmetries, and unpredictable fading. We briefly argue that key ``axioms'' of these types of propagation models lead to simulation results that do not adequately reflect real behavior of ad-hoc networks, and hence to network protocols that may not work well …

A Visual Framework Invites Human Into The Clustering Process, Keke Chen, Ling Liu

Kno.e.sis Publications

Clustering is a technique commonly used in scientific research. The task of clustering inevitably involves human participation - the clustering is not finished when the computer/algorithm finishes but the user has evaluated, understood and accepted the patterns. This defines a human involved "clustering-analysis/evaluation" iteration. Instead of neglecting this human involvement, we provide a visual framework (VISTA) with all power of algorithmic approaches (since their result can be visualized), and in addition we allow the user to steer/monitor/refine the clustering process with domain knowledge. The visual-rendering result also provides a precise pattern for fast post-processing.

A Closer Look At Boosted Image Retrieval, Nicholas Howe

Computer Science: Faculty Publications

Margin-maximizing techniques such as boosting have been generating excitement in machine learning circles for several years now. Although these techniques offer significant improvements over previous methods on classification tasks, little research has examined the application of techniques such as boosting to the problem of retrieval from image and video databases. This paper looks at boosting for image retrieval and classification, with a comparative evaluation of several top algorithms combined in two different ways with boosting. The results show that boosting improves retrieval precision and recall (as expected), but that variations in the way boosting is applied can significantly affect the …

Using Reconstructability Analysis To Select Input Variables For Artificial Neural Networks, Stephen Shervais, Martin Zwick

Complex Systems Faculty Publications and Presentations

We demonstrate the use of Reconstructability Analysis to reduce the number of input variables for a neural network. Using the heart disease dataset we reduce the number of independent variables from 13 to two, while providing results that are statistically indistinguishable from those of NNs using the full variable set. We also demonstrate that rule lookup tables obtained directly from the data for the RA models are almost as effective as NNs trained on model variables.

Level Compressed Dags For Lookup Tables, Ioannis Ioannidis, Ananth Y. Grama

Department of Computer Science Technical Reports

No abstract provided.

Record Linkage: A Machine Learning Approach, A Toolbox, And A Digital Government Web Service, Mohamed G. Elfeky, Vassilios S. Verykios, Ahmed K. Elmagarmid, Thanaa M. Ghanem, Ahmed R. Huwait

Department of Computer Science Technical Reports

No abstract provided.

Distributed Clustering For Scalable, Long-Lived Sensor Networks, Ossama Younis, Sonia Fahmy

Department of Computer Science Technical Reports

No abstract provided.

Violin: Virtual Internetworking On Overlay Infrastructure, Xuxian Jiang, Dongyan Xu

Department of Computer Science Technical Reports

No abstract provided.

Integrating Multiple 3d Views Through Frame-Of-Reference Interaction, Matthew D. Plumlee, Colin Ware

Center for Coastal and Ocean Mapping

Frame-of-reference interaction consists of a unified set of 3D interaction techniques for exploratory navigation of large virtual spaces in nonimmersive environments. It is based on a conceptual framework that considers navigation from a cognitive perspective, as a way of facilitating changes in user attention from one reference frame to another, rather than from the mechanical perspective of moving a camera between different points of interest. All of our techniques link multiple frames of reference in some meaningful way. Some techniques link multiple windows within a zooming environment while others allow seamless changes of user focus between static objects, moving objects, …

Fast And Space-Efficient Location Of Heavy Or Dense Segments In Run-Length Encoded Sequences, Ronald I. Greenberg

Computer Science: Faculty Publications and Other Works

This paper considers several variations of an optimization problem with potential applications in such areas as biomolecular sequence analysis and image processing. Given a sequence of items, each with a weight and a length, the goal is to find a subsequence of consecutive items of optimal value, where value is either total weight or total weight divided by total length. There may also be a specified lower and/or upper bound on the acceptable length of subsequences. This paper shows that all the variations of the problem are solvable in linear time and space even with non-uniform item lengths and divisible …

Investigation Of Third Party Rights Service And Shibboleth Modification To Introduce The Service, Sanket Agrawal

Dartmouth College Master’s Theses

Shibboleth is an architecture to support inter-institutional sharing of electronic resources that are subject to access control. Codifying copyright in Shibboleth authorization policies is difficult because of the copyright exceptions which can be highly subjective. Third Party Rights Service is a high-level concept that has been suggested as a solution to approximate the exceptions of copyright law. In this thesis, I investigate the components of the Third Party Rights Service. I design and analyze a modified Shibboleth architecture based on these components. The resulting architecture allows for the phased addition of the resources to make use of the Third Party …

Multi-Chart Geometry Images, P. V. Sander, Z. J. Wood, S. J. Gortler, H. Hoppe

Computer Science and Software Engineering

We introduce multi-chart geometry images, a new representation for arbitrary surfaces. It is created by resampling a surface onto a regular 2D grid. Whereas the original scheme of Gu et al. maps the entire surface onto a single square, we use an atlas construction to map the surface piecewise onto charts of arbitrary shape. We demonstrate that this added flexibility reduces parametrization distortion and thus provides greater geometric fidelity, particularly for shapes with long extremities, high genus, or disconnected components. Traditional atlas constructions suffer from discontinuous reconstruction across chart boundaries, which in our context create unacceptable surface cracks. Our solution …

A Generative Programming Framework For Adaptive Middleware, Venkita Subramonian, Christopher Gill

All Computer Science and Engineering Research

Historically, many distributed real-time and embedded (DRE) systems were developed manually from scratch, leading to stove-piped solutions that while correct in both functional and QoS properties were very expensive to develop and difficult to maintain and extend. First-generation middleware technologies such as CORBA 2.x [1], XML [2], and SOAP [3], served to shield application developers from low-level platform details, thus raising the level of abstraction at which distributed systems are developed and supporting reuse of infrastructure to amortize development costs over the lifetime of a system. However, interdependencies between services and object interfaces resulting from these programming models significantly limited …

Transformation: The Human Obstacles, Jens G. Pohl

Collaborative Agent Design (CAD) Research Center

This paper addresses the human aspects of change at a time when the rapid wide-spread implementation of a new infostructure has become the cornerstone of national security policies. The author explores the behavioral characteristics of human beings from a point of view that is not based on inter-human comparisons, but on an examination of the innate physical, biological, intellectual, emotional, and social characteristics of our species. It is argued that the experience-based nature of the human cognitive system greatly reduces our ability to adapt to changes in our environment, exploit opportunities, and create new knowledge. A fundamental biological survival instinct …

A Self-Adapting Healthcare Information Infrastructure Using Mobile Computing Devices, James K. Archibald, Eric S. Hall, Charles D. Knutson, David K. Vawdrey

Faculty Publications

Despite recent improvements in the gathering and sharing of patient medical information among healthcare providers, there remains a gap in the electronic medical record infrastructure. Patient data is not available in some situations, either because the infrastructure is inaccessible (as in a natural disaster) or because there is no way to link the patient to the infrastructure (e.g., the patient cannot supply necessary identification information). This paper describes the Poket Doktor System, an architecture that allows an individual to carry personal electronic medical information on a wireless handheld device such as a smart card, cell phone, or PDA. Medical workers …

802.11b Wireless Network Visualization And Radiowave Propagation Modeling, Chris Lentz

Dartmouth College Undergraduate Theses

This paper outlines the methods of creating detailed coverage maps of 802.11b networks, with an emphasis on minimizing the expenses and time involved. The goal of this work is to develop and present a streamlined, reproducible approach to wireless visualization as well as techniques for predicting coverage area before conducting network installations. After evaluating these coverage maps, a repeated series of field measurements will be checked against interpolated values in order to improve techniques for extrapolation of data for unsampled regions. If successful, these extrapolation techniques will provide additional guidelines for, and assist modeling of, new wireless network installations. However, …