Physical Sciences and Mathematics Commons^™

Substituting Failure Avoidance For Redundancy In Storage Fault Tolerance, Christopher David Brumgard

Doctoral Dissertations

The primary mechanism for overcoming faults in modern storage systems is to introduce redundancy in the form of replication and error correcting codes. The costs of such redundancy in hardware, system availability and overall complexity can be substantial, depending on the number and pattern of faults that are handled. This dissertation describes and analyzes, via simulation, a system that seeks to use disk failure avoidance to reduce the need for costly redundancy by using adaptive heuristics that anticipate such failures. While a number of predictive factors can be used, this research focuses on the three leading candidates of SMART errors, …

Applications Of Sampling And Estimation On Networks, Fabricio Murai Ferreira

Doctoral Dissertations

Networks or graphs are fundamental abstractions that allow us to study many important real systems, such as the Web, social networks and scientific collaboration. It is impossible to completely understand these systems and answer fundamental questions related to them without considering the way their components are connected, i.e., their topology. However, topology is not the only relevant aspect of networks. Nodes often have information associated with them, which can be regarded as node attributes or labels. An important problem is then how to characterize a network w.r.t. topology and node label distributions. Another important problem is how to design efficient …

Detecting Anomalously Similar Entities In Unlabeled Data, Lisa D. Friedland

Doctoral Dissertations

In this work, the goal is to detect closely-linked entities within a data set. The entities of interest have a tie causing them to be similar, such as a shared origin or a channel of influence. Given a collection of people or other entities with their attributes or behavior, we identify unusually similar pairs, and we pose the question: Are these two people linked, or can their similarity be explained by chance? Computing similarities is a core operation in many domains, but two constraints differentiate our version of the problem. First, the score assigned to a pair should account for …

Application-Aware Resource Management For Cloud Platforms, Xin He

Doctoral Dissertations

Cloud computing has become increasingly popular in recent years. The benefits of cloud platforms include ease of application deployment, a pay-as-you-go model, and the ability to scale resources up or down based on an application's workload. Today's cloud platforms are being used to host increasingly complex distributed and parallel applications. The main premise of this thesis is that application-aware resource management techniques are better suited for distributed cloud applications over a systems-level one-size-fits-all approach. In this thesis, I study the cloud-based resource management techniques with a particular emphasis on how application-aware approaches can be used to improve system resource utilization …

Elastic Resource Management In Distributed Clouds, Tian Guo

Doctoral Dissertations

The ubiquitous nature of computing devices and their increasing reliance on remote resources have driven and shaped public cloud platforms into unprecedented large-scale, distributed data centers. Concurrently, a plethora of cloud-based applications are experiencing multi-dimensional workload dynamics---workload volumes that vary along both time and space axes and with higher frequency. The interplay of diverse workload characteristics and distributed clouds raises several key challenges for efficiently and dynamically managing server resources. First, current cloud platforms impose certain restrictions that might hinder some resource management tasks. Second, an application-agnostic approach might not entail appropriate performance goals, therefore, requires numerous specific methods. Third, …

A Period Examination Through Contemporary Energy Analysis Of Kevin Roche’S Fine Arts Center At University Of Massachusetts-Amherst, L Carl Fiocchi Jr

Doctoral Dissertations

Studies of buildings belonging to a subset of Modernist architecture, Brutalism, have included discussions pertaining to social and architectural history, critical reception, tectonic form and geometry inspirations, material property selections, period technology limitations, and migration of public perceptions. Evaluations of Brutalist buildings’ energy related performances have been restricted to anecdotal observations with particular focus on the building type’s poor thermal performance, a result of the preferred construction method, i.e. monolithic reinforced concrete used as structure, interior finish and exterior finish. A valid criticism, but one that served to dismiss discussion that the possibility of other positive design strategies limiting energy …

Intrinsic Functions For Securing Cmos Computation: Variability, Modeling And Noise Sensitivity, Xiaolin Xu

Doctoral Dissertations

A basic premise behind modern secure computation is the demand for lightweight cryptographic primitives, like identifier or key generator. From a circuit perspective, the development of cryptographic modules has also been driven by the aggressive scalability of complementary metal-oxide-semiconductor (CMOS) technology. While advancing into nano-meter regime, one significant characteristic of today's CMOS design is the random nature of process variability, which limits the nominal circuit design. With the continuous scaling of CMOS technology, instead of mitigating the physical variability, leveraging such properties becomes a promising way. One of the famous products adhering to this double-edged sword philosophy is the Physically …

Stochastic Network Design: Models And Scalable Algorithms, Xiaojian Wu

Doctoral Dissertations

Many natural and social phenomena occur in networks. Examples include the spread of information, ideas, and opinions through a social network, the propagation of an infectious disease among people, and the spread of species within an interconnected habitat network. The ability to modify a phenomenon towards some desired outcomes has widely recognized benefits to our society and the economy. The outcome of a phenomenon is largely determined by the topology or properties of its underlying network. A decision maker can take management actions to modify a network and, therefore, change the outcome of the phenomenon. A management action is an …

Specification And Analysis Of Resource Utilization Policies For Human-Intensive Systems, Seung Yeob Shin

Doctoral Dissertations

Contemporary systems often require the effective support of many types of resources, each governed by complex utilization policies. Sound management of these resources plays a key role in assuring that these systems achieve their key goals. To help system developers make sound resource management decisions, I provide a resource utilization policy specification and analysis framework for (1) specifying very diverse kinds of resources and their potentially complex resource utilization policies, (2) dynamically evaluating the policies’ effects on the outcomes achieved by systems utilizing the resources, and (3) formally verifying various kinds of properties of these systems. Resource utilization policies range …

Learning From Pairwise Proximity Data, Hamid Dadkhahi

Doctoral Dissertations

In many areas of machine learning, the characterization of the input data is given by a form of proximity measure between data points. Examples of such representations are pairwise differences, pairwise distances, and pairwise comparisons. In this work, we investigate different learning problems on data represented in terms of such pairwise proximities. More specifically, we consider three problems: masking (feature selection) for dimensionality reduction, extension of the dimensionality reduction for time series, and online collaborative filtering. For each of these problems, we start with a form of pairwise proximity which is relevant in the problem at hand. We evaluate the …

Combining Static And Dynamic Analysis For Bug Detection And Program Understanding, Kaituo Li

Doctoral Dissertations

This work proposes new combinations of static and dynamic analysis for bug detection and program understanding. There are 3 related but largely independent directions: a) In the area of dynamic invariant inference, we improve the consistency of dynamically discovered invariants by taking into account second-order constraints that encode knowledge about
invariants; the second-order constraints are either supplied by the programmer or vetted by the programmer (among candidate constraints suggested automatically); b) In the area of testing dataflow (esp. map-reduce) programs, our tool, SEDGE, achieves higher testing coverage by leveraging existing
input data and generalizing them using a symbolic reasoning engine …

Anxiolytic Effects Of Propranolol And Diphenoxylate On Mice And Automated Stretch-Attend Posture Analysis, Kevin Scott Holly

Doctoral Dissertations

The prevention of social anxiety, performance anxiety, and social phobia via the combination of two generic drugs, diphenoxylate HC1 (opioid) plus atropine sulfate (anticholinergic) and propranolol HCl (beta blocker) was evaluated in mice through behavioral studies. A patent published on a September 8, 2011 by Benjamin D. Holly, US 2011/0218215 Al, prompted the research. The drug combination of diphenoxylate and atropine plus propranolol could be an immediate treatment for patients suffering from acute phobic and social anxiety disorders. Demonstrating the anxiolytic effects of the treatment on mice would validate a mouse model for neuroscientist to be used to detect the …

Lattice Boltzmann Methods For Wind Energy Analysis, Stephen Lloyd Wood

Doctoral Dissertations

An estimate of the United States wind potential conducted in 2011 found that the energy available at an altitude of 80 meters is approximately triple the wind energy available 50 meters above ground. In 2012, 43% of all new electricity generation installed in the U.S. (13.1 GW) came from wind power. The majority of this power, 79%, comes from large utility scale turbines that are being manufactured at unprecedented sizes. Existing wind plants operate with a capacity factor of only approximately 30%. Measurements have shown that the turbulent wake of a turbine persists for many rotor diameters, inducing increased vibration …

Achieving High Reliability And Efficiency In Maintaining Large-Scale Storage Systems Through Optimal Resource Provisioning And Data Placement, Lipeng Wan

Doctoral Dissertations

With the explosive increase in the amount of data being generated by various applications, large-scale distributed and parallel storage systems have become common data storage solutions and been widely deployed and utilized in both industry and academia. While these high performance storage systems significantly accelerate the data storage and retrieval, they also bring some critical issues in system maintenance and management. In this dissertation, I propose three methodologies to address three of these critical issues.

First, I develop an optimal resource management and spare provisioning model to minimize the impact brought by component failures and ensure a highly operational experience …

Reducing Power Consumption And Latency In Mobile Devices Using A Push Event Stream Model, Kernel Display Server, And Gui Scheduler, Stephen Gregory Marz

Doctoral Dissertations

The power consumed by mobile devices can be dramatically reduced by improving how mobile operating systems handle events and display management. Currently, mobile operating systems use a pull model that employs a polling loop to constantly ask the operating system if an event exists. This constant querying prevents the CPU from entering a deep sleep, which unnecessarily consumes power.

We’ve improved this process by switching to a push model which we refer to as the event stream model (ESM). This model leverages modern device interrupt controllers which automatically notify an application when events occur, thus removing the need to constantly …

Conditional Computation In Deep And Recurrent Neural Networks, Andrew Scott Davis

Doctoral Dissertations

Recently, deep learning models such as convolutional and recurrent neural networks have displaced state-of-the-art techniques in a variety of application domains. While the computationally heavy process of training is usually conducted on powerful graphics processing units (GPUs) distributed in large computing clusters, the resulting models can still be somewhat heavy, making deployment in resource- constrained environments potentially problematic. In this work, we build upon the idea of conditional computation, where the model is given the capability to learn how to avoid computing parts of the graph. This allows for models where the number of parameters (and in a sense, the …

Quantitative Metrics For Comparison Of Hyper-Dimensional Lsa Spaces For Semantic Differences, John Christopher Martin

Doctoral Dissertations

Latent Semantic Analysis (LSA) is a mathematically based machine learning technology that has demonstrated success in numerous applications in text analytics and natural language processing. The construction of a large hyper-dimensional space, a LSA space, is central to the functioning of this technique, serving to define the relationships between the information items being processed. This hyper-dimensional space serves as a semantic mapping system that represents learned meaning derived from the input content. The meaning represented in an LSA space, and therefore the mappings that are generated and the quality of the results obtained from using the space, is completely dependent …

Efficient Inference, Search And Evaluation For Latent Variable Models Of Text With Applications To Information Retrieval And Machine Translation, Kriste Krstovski

Doctoral Dissertations

Latent variable models of text, such as topic models, have been explored in many areas of natural language processing, information retrieval and machine translation to aid tasks such as exploratory data analysis, automated topic clustering and finding similar documents in mono- and multilingual collections. Many additional applications of these models, however, could be enabled by more efficient techniques for processing large datasets. In this thesis, we introduce novel methods that offer efficient inference, search and evaluation for latent variable models of text. We present efficient, online inference for representing documents in several languages in a common topic space and fast …

Leveraging Backscatter For Ultra-Low Power Wireless Sensing Systems, Pengyu Zhang

Doctoral Dissertations

The past few years have seen a dramatic growth in wireless sensing systems, with millions of wirelessly connected sensors becoming first-class citizens of the Internet. The number of wireless sensing devices is expected to surpass 6.75 billion by 2017, more than the world's population as well as the combined market of smartphones, tablets, and PCs. However, its growth faces two pressing challenges: battery energy density and wireless radio power consumption. Battery energy density looms as a fundamental limiting factor due to slow improvements over the past several decades (3x over 22 years). Wireless radio power consumption is another key challenge …

Wind Farm Wake Modeling And Analysis Of Wake Impacts In A Wind Farm, Yujia Hao

Doctoral Dissertations

More and more wind turbines have been grouped in the same location during the last decades to take the advantage of profitable wind resources and reduced maintenance cost. However wind turbines located in a wind farm are subject to a wind field that is substantially modified compared to the ambient wind field due to wake effects. The wake results in a reduced power production, increased load variation on the waked turbine, and reduced wake farm efficiency. Therefore the wake has long been an important concern for the wind farm installation, maintenance, and control. Thus a wake simulation tool is required. …

Extending Faceted Search To The Open-Domain Web, Weize Kong

Doctoral Dissertations

Faceted search enables users to navigate a multi-dimensional information space by combining keyword search with drill-down options in each facets. For example, when searching “computer monitor”' in an e-commerce site, users can select brands and monitor types from the the provided facets {“Samsung”, “Dell”, “Acer”, ...} and {“LET-Lit”, “LCD”, “OLED”, ...}. It has been used successfully for many vertical applications, including e-commerce and digital libraries. However, this idea is not well explored for general web search in an open-domain setting, even though it holds great potential for assisting multi-faceted queries and exploratory search. The goal of this work is to …

Effective Performance Analysis And Debugging, Charles M. Curtsinger

Doctoral Dissertations

Performance is once again a first-class concern. Developers can no longer wait for the next generation of processors to automatically "optimize" their software. Unfortunately, existing techniques for performance analysis and debugging cannot cope with complex modern hardware, concurrent software, or latency-sensitive software services. While processor speeds have remained constant, increasing transistor counts have allowed architects to increase processor complexity. This complexity often improves performance, but the benefits can be brittle; small changes to a program’s code, inputs, or execution environment can dramatically change performance, resulting in unpredictable performance in deployed software and complicating performance evaluation and debugging. Developers seeking to …

Direct And Inverse Scattering Problems For Domains With Multiple Corners, Jiang Yihong

Doctoral Dissertations

Direct and inverse scattering problems have wide applications in geographical exploration, radar, sonar, medical imaging and non-destructive testing. In many applications, the obstacles are not smooth. Corner singularity challenges the design of a forward solver. Also, the nonlinearity and ill-posedness of the inverse problem challenge the design of an efficient, robust and accurate imaging method.

This dissertation presents numerical methods for solving the direct and inverse scattering problems for domains with multiple corners. The acoustic wave is sent from the transducers, scattered by obstacles and received by the transducers. This forms the response matrix data. The goal for the direct …

An Intelligent Robot And Augmented Reality Instruction System, Christopher M. Reardon

Doctoral Dissertations

Human-Centered Robotics (HCR) is a research area that focuses on how robots can empower people to live safer, simpler, and more independent lives. In this dissertation, I present a combination of two technologies to deliver human-centric solutions to an important population. The first nascent area that I investigate is the creation of an Intelligent Robot Instructor (IRI) as a learning and instruction tool for human pupils. The second technology is the use of augmented reality (AR) to create an Augmented Reality Instruction (ARI) system to provide instruction via a wearable interface.

To function in an intelligent and context-aware manner, both …

Performance Analysis And Modeling Of Task-Based Runtimes, Blake Andrew Haugen

Doctoral Dissertations

The shift toward multicore processors has transformed the software and hardware landscape in the last decade. As a result, software developers must adopt parallelism in order to efficiently make use of multicore CPUs. Task-based scheduling has emerged as one method to reduce the complexity of parallel computing. Although task-based scheduling has been around for many years, the inclusion of task dependencies in OpenMP 4.0 suggests the paradigm will be around for the foreseeable future.

While task-based schedulers simplify the process of parallel software development, they can obfuscate the performance characteristics of the execution of an algorithm. Additionally, they can create …

Shape Design And Optimization For 3d Printing, Yahan Zhou

Doctoral Dissertations

In recent years, the 3D printing technology has become increasingly popular, with wide-spread uses in rapid prototyping, design, art, education, medical applications, food and fashion industries. It enables distributed manufacturing, allowing users to easily produce customized 3D objects in office or at home. The investment in 3D printing technology continues to drive down the cost of 3D printers, making them more affordable to consumers. As 3D printing becomes more available, it also demands better computer algorithms to assist users in quickly and easily generating 3D content for printing. Creating 3D content often requires considerably more efforts and skills than creating …

Algorithms For First-Order Sparse Reinforcement Learning, Bo Liu

Doctoral Dissertations

This thesis presents a general framework for first-order temporal difference learning algorithms with an in-depth theoretical analysis. The main contribution of the thesis is the development and design of a family of first-order regularized temporal-difference (TD) algorithms using stochastic approximation and stochastic optimization. To scale up TD algorithms to large-scale problems, we use first-order optimization to explore regularized TD methods using linear value function approximation. Previous regularized TD methods often use matrix inversion, which requires cubic time and quadratic memory complexity. We propose two algorithms, sparse-Q and RO-TD, for on-policy and off-policy learning, respectively. These two algorithms exhibit linear computational …

Algorithms Leveraging Smartphone Sensing For Analyzing Explosion Events, Srinivas Chakravarthi Thandu

Doctoral Dissertations

"The increasing frequency of explosive disasters throughout the world in recent years have created a clear need for the systems to monitor for them continuously to improve the post-disaster emergency events such as rescue and recovery operations. Disasters both man-made and natural are unfortunate and not preferred, however monitoring them may be a lifesaving phenomenon in emergency scenarios. Dedicated sensors deployed in the public places and their associated networks to monitor such events may be inadequate and must be complemented for making the monitoring more pervasive and effective. In the recent past, modern smartphones with significant processing, networking and storage …

Mechanisms For Improving Information Quality In Smartphone Crowdsensing Systems, Francesco Restuccia

Doctoral Dissertations

"Given its potential for a large variety of real-life applications, smartphone crowdsensing has recently gained tremendous attention from the research community. Smartphone crowdsensing is a paradigm that allows ordinary citizens to participate in large-scale sensing surveys by using user-friendly applications installed in their smartphones. In this way, fine-grained sensing information is obtained from smartphone users without employing fixed and expensive infrastructure, and with negligible maintenance costs.

Existing smartphone sensing systems depend completely on the participants' willingness to submit up-to-date and accurate information regarding the events being monitored. Therefore, it becomes paramount to scalably and effectively determine, enforce, and optimize the …

Models Of Leader Elections And Their Applications, Stephen Curtis Jackson

Doctoral Dissertations

"New research about cyber-physical systems is rapidly changing the way we think about critical infrastructures such as the power grid. Changing requirements for the generation, storage, and availability of power are all driving the development of the smart-grid. Many smart-grid projects disperse power generation across a wide area and control devices with a distributed system. However, in a distributed system, the state of processes is hard to determine due to isolation of memory. By using information flow security models, we reason about a process's beliefs of the system state in a distributed system. Information flow analysis aided in the creation …