UMD Theses and Dissertations
Permanent URI for this collectionhttp://hdl.handle.net/1903/3
New submissions to the thesis/dissertation collections are added automatically as they are received from the Graduate School. Currently, the Graduate School deposits all theses and dissertations from a given semester after the official graduation date. This means that there may be up to a 4 month delay in the appearance of a given thesis/dissertation in DRUM.
More information is available at Theses and Dissertations at University of Maryland Libraries.
Browse
9 results
Search Results
Item SOCIAL CATALYSTS AND SOCIAL GOAL PURSUIT(2020) Irions, Amanda L; Hample, Dale J; Communication; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)People pursue goals but do not always successfully attain them. Existing theories of goal pursuit such as field theory and the goals-plans-actions model regard goal pursuit as a solitary activity that results either in success or frustrated failure. In stark contrast to this solitary-actor, sink-or-swim model of goal pursuit are observations from several social domains show that people ask other people to help them reach their goals instead of abandoning their goals entirely. This dissertation presents the quantitative findings from two studies of these helpers, and argues that analyzing and developing a theory of helpers is critical to a more complete and accurate model of goal pursuit. By introducing the constructs of resource improvement (helpers increase resources, diversify resources, and show their pursuers new paths around obstacles blocking goal pursuit) and the substitutability of helpers’ willingness and skills, this dissertation demonstrates the utility of unifying goal-pursuit theories with the social-support framework and situating those ideas in a social context. Study 1 reports an investigation of wingpeople, those offensive and defensive helpers (also called wingmen) who use communication to help people initiate or terminate initial romantic relationships. Key findings include that both offensive and defensive wingpeople use communication to help pursuers move toward a desired potential romantic partner and away from an undesirable one and that, in line with evolutionary psychological predications, wingpeople provided differential help to male and female pursuers. Notably, some participants in Study 1 spontaneously reported being helpers in social domains other than courtship. Study 2 investigated the generalizability of the helping phenomenon across social domains. Key findings include: participants reported being helpers in more than a dozen different social domains (e.g., academic, physical health, creative pursuits, and service) and more than 90% reported helping in domains other than courtship; participants used social support messages to improve their pursuers’ resources; and no differences between offensive and defensive helpers were observed on the personality traits measured. This dissertation concludes by using the evidence from the studies to make a case for helpers as social catalysts.Item Experiments on networks of coupled opto-electronic oscillators and physical random number generators(2018) Hart, Joseph David; Roy, Rajarshi; Murphy, Thomas E; Physics; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)In this thesis, we report work in two areas: synchronization in networks of coupled oscillators and the evaluation of physical random number generators. A ``chimera state'' is a dynamical pattern that occurs in a network of coupled identical oscillators when the symmetry of the oscillator population is spontaneously broken into coherent and incoherent parts. We report a study of chimera states and cluster synchronization in two different opto-electronic experiments. The first experiment is a traditional network of four opto-electronic oscillators coupled by optical fibers. We show that the stability of the observed chimera state can be determined using the same group-theoretical techniques recently developed for the study of cluster synchrony. We present three novel results: (i) chimera states can be experimentally observed in small networks, (ii) chimera states can be stable, and (iii) at least some types of chimera states (those with identically synchronized coherent regions) are closely related to cluster synchronization. The second experiment uses a single opto-electronic feedback loop to investigate the dynamics of oscillators coupled in large complex networks with arbitrary topology. Recent work has demonstrated that an opto-electronic feedback loop can be used to realize ring networks of coupled oscillators. We significantly extend these capabilities and implement networks with arbitrary topologies by using field programmable gate arrays (FPGAs) to design appropriate digital filters and time delays. With this system, we study (i) chimeras in a five-node globally-coupled network, (ii) synchronization of clusters that are not predicted by network symmetries, and (iii) optimal networks for cluster synchronization. The field of random number generation is currently undergoing a fundamental shift from relying solely on pseudo-random algorithms to employing physical entropy sources. The standard evaluation practices, which were designed for pseudo-random number generators, are ill-suited to quantify the entropy that underlies physical random number generation. We review the state of the art in the evaluation of physical random number generation and recommend a new paradigm: quantifying entropy generation and understanding the physical limits for harvesting entropy from sources of randomness. As an illustration of our recommendations, we evaluate three common optical entropy sources: single photon time-of-arrival detection, chaotic lasers, and amplified spontaneous emission.Item Optimality of Event-Based Policies for Decentralized Estimation over Shared Networks(2016) Vasconcelos, Marcos Muller; Martins, Nuno C; Electrical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Cyber-physical systems often consist of multiple non-collocated components that sense, exchange information and act as a team through a network. Although this new paradigm provides convenience, flexibility and robustness to modern systems, design methods to achieve optimal performance are elusive as they must account for certain detrimental characteristics of the underlying network. These include constrained connectivity among agents, rate-limited communication links, physical noise at the antennas, packet drops and interference. We propose a new class of problems in optimal networked estimation where multiple sensors operating as a team communicate their measurements to a fusion center over an interference prone network modeled by a collision channel. Using a team decision theoretic approach, we characterize jointly optimal communication policies for one-shot problems under different performance criteria. First we study the problem of estimating two independent continuous random variables observed by two different sensors communicating with a fusion center over a collision channel. For a minimum mean squared estimation error criterion, we show that there exist team-optimal strategies where each sensor uses a threshold policy. This result is independent of the distribution of the observations and, can be extended to vector observations and to any number of sensors. Consequently, the existence of team-optimal threshold policies is a result of practical significance, because it can be applied to a wide class of systems without requiring collision avoidance protocols. Next we study the problem of estimating independent discrete random variables over a collision channel. Using two different criteria involving the probability of estimation error, we show the existence of team-optimal strategies where the sensors either transmit all but the most likely observation; transmit only the second most likely observation; or remain always silent. These results are also independent of the distributions and are valid for any number of sensors. In our analysis, the proof of the structural result involves the minimization of a concave functional, which is an evidence of the inherent complexity of team decision problems with nonclassical information structure. In the last part of the dissertation, the assumption on the cooperation among sensors is relaxed, and we show that similar structural results can also be obtained for systems with one or more selfish sensors. Finally the assumption of the independence is lifted by introducing the observation of a common random variable in addition to the private observations of each sensor. The structural result obtained provides valuable insights on the characterization of team-optimal policies for a general correlation structure between the observed random variables.Item CROSS-SECTOR COLLABORATIVES: ISSUES OF IMPLEMENTATION AND PERFORMANCE(2014) Littlefield, Jennifer Nash; Kettl, Donald; Grimm, Robert; Public Policy; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Poverty, access to quality education, unemployment... all examples of complex issues that demand attention in our society. For these problems, solutions are often forged through joint action in the form of collaborative networks. Collaborative networks are defined as "collections of government agencies, nonprofits, and for-profits that work together to provide a public good, service, or `value' when a single agency is unable to create the good or service on its own (Isett, Mergel, LeRoux, Mischen, and Rethemeyer, 2011, p. i158)." This dissertation examines the relationship between the collaborative design and implementation process and collaborative effectiveness. I include a comparative case study method and utilize the multiple-case replication design (Yin, 2009); specifically analyzing six cases from the Annie E. Casey Leadership in Action Program (LAP). Interviews, document analysis and an original survey are used as part of the research design. This dissertation has two key components. First, I operationalize and expand an important evaluative tool that allows collaboratives to understand their performance at various levels and share their success and shortcomings in a rich, straightforward, and cost effective manner. This framework allows for measurement on multiple dimensions and levels, lending information on the relevance and impact of collaborative groups. Secondly, I use my findings with regard to performance to analyze the process of high, moderate, and low performing groups to determine the most important elements of successful collaboration. This research demonstrates a clear relationship between design process and effectiveness, with certain elements making positive results more likely. These are: the use of an accountability system, decision-making process, relationship building, and facilitation. Overall this research fills a void and makes a significant contribution to the literature and practice of collaborative networks, potentially impacting how future cross-sector collaboratives work together to produce public value and address major public problems.Item Authority Flow-based Ranking in Heterogeneous Networks: Prediction, Personalization, and Learning to Rank(2014) Sayyadi Harikandehei, Hassan; Raschid, Louiqa; Computer Science; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Many real-world datasets, including biological networks, the Web, and social media, can be effectively modeled as networks or graphs, in which nodes represent entities of interest and links mimic the interactions or relationships among them. Such networks often contain multiple entity or relationship types, which are referred to as heterogeneous networks. Networks also evolve due to the existence of temporal features that characterize the entities or to the temporal relationships among them. Finding important/authoritative entities in real-world networks is a long-standing and well-defined challenge. In this dissertation, I focus on two variants of the problem. The first is the prediction of the ranking of scientific publications in a future state of a citation network. I introduce a new measure labeled the future PageRank score. I develop FutureRank, a prediction algorithm for predicting the future PageRank scores from the historical network structure, and evaluate the FutureRank algorithm on multiple bibliographic dataset. Next, I focus on personalized ranking in social media. I extend a social media dataset to include relationships (edge types) between authors, blog posts, categories (topics) of the posts, and events (collections of posts). I then apply personalized ranking algorithms over the historical posts and events that have been visited by a user and use the ranking to recommend additional posts. I evaluate the personalized recommendations through an experiment with real users, as well as an extensive study of synthetic users whose preferences are defined based on intuitive criteria. Finally, I present an approach for learning to rank (algorithms) applied to heterogeneous networks. Existing methods for learning to rank are typically limited to content-based features, while many real world problems correspond to relational features. I develop a framework for learning to rank, which targets authority flow-based ranking models on heterogeneous networks. I propose algorithms for both pointwise and pairwise learning. However, this framework can easily utilize any loss function from a non-relational learning domain. Experiments show that even with a small amount of training data, both pointwise and pairwise algorithms perform successfully and converge very fast. In addition, these solutions are shown to be robust against noise.Item Marketing Applications of Social Tagging Networks(2012) Nam, Hyoryung; Kannan, P.K.; Joshi, Yogesh; Business and Management: Marketing; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)This dissertation focuses on marketing applications of social tagging networks. Social tagging is a new way to share and categorize content, allowing users to express their perceptions and feelings with respect to concepts such as brands and firms with their own keywords, “tags.” The associative information in social tagging networks provides marketers with a rich source of information reflecting consumers’ mental representations of a brand/firm/product. The first essay presents a methodology to create “social tag maps,” brand associative networks derived from social tags. The proposed approach reflects a significant improvement towards understanding brand associations compared to conventional techniques (e.g., brand concept maps and recent text mining techniques), and helps marketers to track real-time updates in a brand’s associative network and dynamically visualize the relative competitive position of their brand. The second essay investigates how information contained in social tags acts as proxy measures of brand assets that track and predict the financial valuation of firms using the data collected from a social bookmarking website, del.icio.us, for 61 firms across 16 industries. The results suggest that brand asset metrics based on social tags explain stock return. Specifically, an increase in social attention and connectedness to competitors is shown to be positively related to stock return for less prominent brands, while for prominent brands associative uniqueness and evaluation valence is found to be more significantly related to stock return. The findings suggest to marketing practitioners a new way to proactively improve brand assets for impacting a firm’s financial performance. The third essay investigates whether the position of products on social tagging networks can predict sales dynamics. We find that (1) books in long tail can increase sales by being strongly linked to well-known keywords with high degree centrality and (2) top sellers can be better sellers by creating dense content clusters rather than connecting them to well-known keywords with high degree centrality. Our findings suggest that marketing managers better understand a user community’s perception of products and potentially influence product sales by taking into account the positioning of their products within social tagging networks.Item Reliability Evaluation of Common-Cause Failures and Other Interdependencies in Large Reconfigurable Networks(2010) Guenzi, Giancarlo; Mosleh, Ali; Reliability Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)This work covers the impact of Interdependencies and CCFs in large repairable networks with possibility of "re-configuration" after a fault and the consequent disconnection of the faulted equipment. Typical networks with these characteristics are the Utilities, e.g. Power Transmission and Distribution Systems, Telecommunication Systems, Gas and Water Utilities, Wi Fi networks. The main issues of the research are: (a) Identification of the specific interdependencies and CCFs in large repairable networks, and (b)Evaluation of their impact on the reliability parameters (load nodes availability, etc.). The research has identified (1) the system and equipment failure modes that are relevant to interdependencies and CCF, and their subsequent effects, and (2) The hidden interdependencies and CCFs relevant to control, supervision and protection systems, and to the automatic change-over systems, that have no impact in normal operation, but that can cause relevant out-of-service when the above automatic systems are called to operate under and after fault conditions. Additionally methods were introduced to include interdependencies and CCFs in the reliability and availability models. The results of the research include a new generalized approach to model the repairable networks for reliability analysis, including Interdependencies/CCFs as a main contributor. The method covers Generalized models for Nodes, Branches and Load nodes; Interdependencies and CCFs on Networks / Components; System Interdependencies/CCFs; Functional Interdependencies/CCFs; Simultaneous and non-simultaneous Interdependencies/CCFs. As an example detailed Interdependency/CCFs analysis and generalized model of an important network structure (a "RING" with load nodes) has been analyzed in detail.Item DISTRIBUTED ESTIMATION OVER NETWORKS WITH COMMUNICATION COSTS(2010) Lipsa, Gabriel-Mihai; Martins, Nuno C; Electrical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)We analyze how distributed or decentralized estimation can be performed over networks, when there is a price to be paid whenever nodes in the network communicate with each other. The work here has application especially in the network control systems. Assume that different nodes in the network can track perfectly or with imperfectly some stochastic processes, while other nodes in the network need to estimate these stochastic processes. The nodes which can observe the stochastic processes can send information directly to the nodes which need to estimate the processes, or information can be sent to intermediate nodes. When each transmission is performed a cost for communication is paid. The goal of the network is to optimize jointly a cost which consists both of a function of the estimation error and a function of the transmission cost. We show here that for some simple topologies the decision to send information over the network is a threshold policy, while the estimators are linear estimators which resemble with the Kalman-filter. For the result dealing with simple topologies we have proved the results using majorization theory. It is also shown here both analytically and numerically that things can immediately become quite complicated. If we take into consideration multidimensional problems or problems with multiple agents and/or transmission noise, the optimal strategies can no longer be found analytically and it can be quite difficult to compute numerically the optimal strategies.Item Friends and Partners: The Impact of Network Ties(2009) Cangiano, Giulia Cristina; Murrell, Peter; Kranton, Rachel; Economics; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)How does a high-tech entrepreneur find the most qualified engineer for her startup? How does a scientific inventor acquire funding or recruit the best partner for his project? In chapter 1 I develop a discrete matching model with heterogeneous values and an undirected social network to address these questions. My model offers a framework to study how relative network positions affect payoffs and incentives. While an entrepreneur's expected return increases with the size of her own network, the network externalities from competing entrepreneurs are more complex. There is a tradeoff between the size of an entrepreneur's network and the competitive externality she exerts. When an entrepreneur's network increases, her closest competitors are hurt, but her less similar competitors may actually have a better chance of finding a suitable partner. In a more connected network, fewer frictions interfere with compatible matches. Results are consistent with observable patterns in high-tech and biotechnology in Silicon Valley and Massachusetts, as well as the turn of the 20th century German synthetic dye manufacturing. Initiatives to promote social networks within innovative sectors are critical and deserve future research. In Chapter 2 I consider a two-period endogenous network search model in which entrepreneurs build relationships with specialists. The model includes a period of costly network search and applies results from my companion paper. In the presence of network externalities, entrepreneurs over-invest in networking. Networks in which is it not costly to build new relationships are the least efficient. While positive externalities reduce this problem some negative inefficiencies will likely prevail. Networks in which participation is cheap - such as online career networks LinkedIn or Monster.com - have limited information about individual specialists and are the most inefficient. A network that is costly to participate in, but is more effective at targeting entrepreneur's search for qualified candidates results in a more compatible and, likely, efficient partnership. These networks might include alumni groups, trade associations or head-hunters. This chapter provides one explanation for the varied successes of government programs in fostering effective business networks. Efficient networks foster fewer, more specific relationships.