ASU Electronic Theses and Dissertations
This collection includes most of the ASU Theses and Dissertations from 2011 to present. ASU Theses and Dissertations are available in downloadable PDF format; however, a small percentage of items are under embargo. Information about the dissertations/theses includes degree information, committee members, an abstract, supporting data or media.
In addition to the electronic theses found in the ASU Digital Repository, ASU Theses and Dissertations can be found in the ASU Library Catalog.
Dissertations and Theses granted by Arizona State University are archived and made available through a joint effort of the ASU Graduate College and the ASU Libraries. For more information or questions about this collection contact or visit the Digital Repository ETD Library Guide or contact the ASU Graduate College at gradformat@asu.edu.
Displaying 1 - 10 of 299
Filtering by
- All Subjects: Machine Learning
Description
I study the performance of hedge fund managers, using quarterly stock holdings from 1995 to 2010. I use the holdings-based measure built on Ferson and Mo (2012) to decompose a manager's overall performance into stock selection and three components of timing ability: market return, volatility, and liquidity. At the aggregate level, I find that hedge fund managers have stock picking skills but no timing skills, and overall I do not find strong evidence to support their superiority. I show that the lack of abilities is driven by the large fluctuations of timing performance with market conditions. I find that conditioning information, equity capital constraints, and priority in stocks to liquidate can partly explain the weak evidence. At the individual fund level, bootstrap analysis results suggest that even top managers' abilities cannot be separated from luck. Also, I find that hedge fund managers exhibit short-horizon persistence in selectivity skill.
ContributorsKang, MinJeong (Author) / Aragon, George O. (Thesis advisor) / Hertzel, Michael G (Committee member) / Boguth, Oliver (Committee member) / Arizona State University (Publisher)
Created2013
Description
Sparsity has become an important modeling tool in areas such as genetics, signal and audio processing, medical image processing, etc. Via the penalization of l-1 norm based regularization, the structured sparse learning algorithms can produce highly accurate models while imposing various predefined structures on the data, such as feature groups or graphs. In this thesis, I first propose to solve a sparse learning model with a general group structure, where the predefined groups may overlap with each other. Then, I present three real world applications which can benefit from the group structured sparse learning technique. In the first application, I study the Alzheimer's Disease diagnosis problem using multi-modality neuroimaging data. In this dataset, not every subject has all data sources available, exhibiting an unique and challenging block-wise missing pattern. In the second application, I study the automatic annotation and retrieval of fruit-fly gene expression pattern images. Combined with the spatial information, sparse learning techniques can be used to construct effective representation of the expression images. In the third application, I present a new computational approach to annotate developmental stage for Drosophila embryos in the gene expression images. In addition, it provides a stage score that enables one to more finely annotate each embryo so that they are divided into early and late periods of development within standard stage demarcations. Stage scores help us to illuminate global gene activities and changes much better, and more refined stage annotations improve our ability to better interpret results when expression pattern matches are discovered between genes.
ContributorsYuan, Lei (Author) / Ye, Jieping (Thesis advisor) / Wang, Yalin (Committee member) / Xue, Guoliang (Committee member) / Kumar, Sudhir (Committee member) / Arizona State University (Publisher)
Created2013
Description
This work presents two complementary studies that propose heuristic methods to capture characteristics of data using the ensemble learning method of random forest. The first study is motivated by the problem in education of determining teacher effectiveness in student achievement. Value-added models (VAMs), constructed as linear mixed models, use students’ test scores as outcome variables and teachers’ contributions as random effects to ascribe changes in student performance to the teachers who have taught them. The VAMs teacher score is the empirical best linear unbiased predictor (EBLUP). This approach is limited by the adequacy of the assumed model specification with respect to the unknown underlying model. In that regard, this study proposes alternative ways to rank teacher effects that are not dependent on a given model by introducing two variable importance measures (VIMs), the node-proportion and the covariate-proportion. These VIMs are novel because they take into account the final configuration of the terminal nodes in the constitutive trees in a random forest. In a simulation study, under a variety of conditions, true rankings of teacher effects are compared with estimated rankings obtained using three sources: the newly proposed VIMs, existing VIMs, and EBLUPs from the assumed linear model specification. The newly proposed VIMs outperform all others in various scenarios where the model was misspecified. The second study develops two novel interaction measures. These measures could be used within but are not restricted to the VAM framework. The distribution-based measure is constructed to identify interactions in a general setting where a model specification is not assumed in advance. In turn, the mean-based measure is built to estimate interactions when the model specification is assumed to be linear. Both measures are unique in their construction; they take into account not only the outcome values, but also the internal structure of the trees in a random forest. In a separate simulation study, under a variety of conditions, the proposed measures are found to identify and estimate second-order interactions.
ContributorsValdivia, Arturo (Author) / Eubank, Randall (Thesis advisor) / Young, Dennis (Committee member) / Reiser, Mark R. (Committee member) / Kao, Ming-Hung (Committee member) / Broatch, Jennifer (Committee member) / Arizona State University (Publisher)
Created2013
Description
"YouTube Shakespeares" is a study of Shakespeare online videos and the people who create, upload, and view them on YouTube. Employing an interdisciplinary approach, this work is a remix of theories and methodologies from literary, performance, (social) media, fan, and Internet studies that expands the field of Shakespeare studies. This dissertation explores the role of YouTube users and their activities, the expansion of literary research methods onto digital media venues, YouTube as site of Shakespeare performance, and YouTube Shakespeares' fan communities. It analyzes a broad array of Shakespeare visual performances including professional and user-generated mashups, remixes, film clips, auditions, and high school performances. A rich avenue for the study of people's viewing and reception of Shakespeare, YouTube tests the (un)limitations of Shakespeare adaptation. This work explores the ethical implications of researching performances that include human subjects, arguing that their presence frequently complicates common concepts of public and private identities. Although YouTube is a "published" forum for social interactivity and video repository, this work urges digital humanities scholars to recognize and honor the human users entailed in the videos not as text, but as human subjects. Shifting the study focus to human subjects demands a revision of research methods and publications protocols as the researcher repositions herself into the role of virtual ethnographer. "YouTube Shakespeares" develops its own ethics-based, online research method, which includes seeking Institutional Board Review approval and online interviews. The second half of the dissertation shifts from methodology to theorizing YouTube Shakespeares' performance spaces as analogs to the interactive and imaginary areas of Shakespeare's early modern theatre. Additionally, this work argues that YouTube Shakespeares' creators and commentators are fans. "YouTube Shakespeares" is one of the first Shakespeare-centric studies to employ fan studies as a critical lens to explore the cultural significance and etiquette of people's online Shakespeare performance activities. The work ends with a conversation about the issues of ephemerality, obsolescence, and concerns about the instability of digital and online materials, noting the risk of evidentiary loss of research materials is far outweighed by a scholarly critical registration of YouTube in the genealogy of Shakespeare performance.
ContributorsFazel, Valerie Margaret (Author) / Thompson, Ayanna (Thesis advisor) / Ryner, Bradley (Committee member) / Fox, Cora (Committee member) / Arizona State University (Publisher)
Created2013
Description
The integration of yoga into the music curriculum has the potential of offering many immediate and life-long benefits to musicians. Yoga can help address issues such as performance anxiety and musculoskeletal problems, and enhance focus and awareness during musical practice and performance. Although the philosophy of yoga has many similarities to the process of learning a musical instrument, the benefits of yoga for musicians is a topic that has gained attention only recently. This document explores several ways in which the practice and philosophy of yoga can be fused with saxophone pedagogy as one way to prepare students for a healthy and successful musical career. A six-week study at Arizona State University was conducted to observe the effects of regular yoga practice on collegiate saxophone students. Nine participants attended a sixty-minute "yoga for musicians" class twice a week. Measures included pre- and post- study questionnaires as well as personal journals kept throughout the duration of the study. These self-reported results showed that yoga had positive effects on saxophone playing. It significantly increased physical comfort and positive thinking, and improved awareness of habitual patterns and breath control. Student participants responded positively to the idea of integrating such a course into the music curriculum. The integration of yoga and saxophone by qualified professionals could also be a natural part of studio class and individual instruction. Carrie Koffman, professor of saxophone at The Hartt School, University of Hartford, has established one strong model for the combination of these disciplines. Her methods and philosophy, together with the basics of Western-style hatha yoga, clinical reports on performance injuries, and qualitative data from the ASU study are explored. These inquiries form the foundation of a new model for integrating yoga practice regularly into the saxophone studio.
ContributorsAdams, Allison Dromgold (Author) / Norton, Kay (Thesis advisor) / Hill, Gary (Committee member) / McAllister, Timothy (Committee member) / Micklich, Albie (Committee member) / Standley, Eileen (Committee member) / Arizona State University (Publisher)
Created2012
Description
The Holocaust and the effects it has had upon witnesses has been a topic of study for nearly six decades; however, few angles of research have been conducted relating to the long-term effects of the Holocaust upon the children and grandchildren of Holocaust survivors--the After Generations. The After Generations are considered the proof--the living legacies--that their parents and grandparents survived. Growing up with intimate knowledge of the atrocities that occurred during the Holocaust, members of the After Generations not only carry with them their family's story, but also their own vicarious experience(s) of trauma. From this legacy comes a burden of responsibility to those who perished, their survivor parents/grandparents, the stories that were shared, as well as to future generations. Using grounded theory method, this study not only explores the long-term effects of the Holocaust upon members of the After Generations, but what it means to responsibly remember the stories from the Holocaust, as well as how individuals might ethically represent such stories/memories. Findings that developed out of an axial analysis of interview transcripts and journal writing, as well as the later development of a performance script, are embodied in a manner that allows the actual language and experiences of the participants to be collectively witnessed both symbolically and visually. Through their desire to remember, members of the After Generations demonstrate how they plan to carry on traditions, live lives that honor those that came before them, and maintain hope for the future. In so doing, the stories shared reveal the centrality of the Holocaust in the lives of members of the After Generations through their everyday choices to responsibly and actively remember through their art, writings, life-work, as well as from within their work in their local communities. Such acts of remembrance are important to the education of others as well as to the construction and maintenance of the After Generations' identities. The representation of these voices acts as a reminder of how hatred and its all-consuming characteristics can affect not only the person targeted, but multiple generations, as well.
ContributorsRath, Sandra (Author) / de la Garza, Sarah Amira (Thesis advisor) / Underiner, Tamara (Committee member) / Corey, Frederick C. (Committee member) / Eisenberg, Judith (Committee member) / Arizona State University (Publisher)
Created2012
Description
Image understanding has been playing an increasingly crucial role in vision applications. Sparse models form an important component in image understanding, since the statistics of natural images reveal the presence of sparse structure. Sparse methods lead to parsimonious models, in addition to being efficient for large scale learning. In sparse modeling, data is represented as a sparse linear combination of atoms from a "dictionary" matrix. This dissertation focuses on understanding different aspects of sparse learning, thereby enhancing the use of sparse methods by incorporating tools from machine learning. With the growing need to adapt models for large scale data, it is important to design dictionaries that can model the entire data space and not just the samples considered. By exploiting the relation of dictionary learning to 1-D subspace clustering, a multilevel dictionary learning algorithm is developed, and it is shown to outperform conventional sparse models in compressed recovery, and image denoising. Theoretical aspects of learning such as algorithmic stability and generalization are considered, and ensemble learning is incorporated for effective large scale learning. In addition to building strategies for efficiently implementing 1-D subspace clustering, a discriminative clustering approach is designed to estimate the unknown mixing process in blind source separation. By exploiting the non-linear relation between the image descriptors, and allowing the use of multiple features, sparse methods can be made more effective in recognition problems. The idea of multiple kernel sparse representations is developed, and algorithms for learning dictionaries in the feature space are presented. Using object recognition experiments on standard datasets it is shown that the proposed approaches outperform other sparse coding-based recognition frameworks. Furthermore, a segmentation technique based on multiple kernel sparse representations is developed, and successfully applied for automated brain tumor identification. Using sparse codes to define the relation between data samples can lead to a more robust graph embedding for unsupervised clustering. By performing discriminative embedding using sparse coding-based graphs, an algorithm for measuring the glomerular number in kidney MRI images is developed. Finally, approaches to build dictionaries for local sparse coding of image descriptors are presented, and applied to object recognition and image retrieval.
ContributorsJayaraman Thiagarajan, Jayaraman (Author) / Spanias, Andreas (Thesis advisor) / Frakes, David (Committee member) / Tepedelenlioğlu, Cihan (Committee member) / Turaga, Pavan (Committee member) / Arizona State University (Publisher)
Created2013
Description
Research has shown that the ability to smell is the most direct sense an individual can experience. With every breath a person takes, the brain recognizes thousands of molecules and makes connections with our memories to determine their composition. With the amount of research looking into how and why we smell, researchers still have little understanding of how the nose and brain process an aroma, and how emotional and physical behavior is impacted. This research focused on the affects smell has on a caregiver in a simulated Emergency Department setting located in the SimET of Banner Good Samaritan Medical Center in Phoenix, Arizona. The study asked each participant to care for a programmed mannequin, or "patient", while performing simple computer-based tasks, including memory and recall, multi-tasking, and mood-mapping to gauge physical and mental performance. Three different aromatic environments were then introduced through diffusion and indirect inhalation near the participants' task space: 1) a control (no smell), 2) an odor (simulated dirty feet), and 3) an aroma (one of four true essential oils plus a current odor-eliminating compound used in many U.S. Emergency Departments). This study was meant to produce a stressful environment by leading the caregiver to stay in constant movement throughout the study through timed tasks, uncooperative equipment, and a needy "patient". The goal of this research was to determine if smells, and of what form of pleasantness and repulsiveness, can have an effect on the physical and mental performance of emergency caregivers. Findings from this study indicated that the "odor eliminating" method currently used in typical Emergency Departments, coffee grounds, is more problematic than helpful, and the introduction of true essential oils may not only reduce stress, but increase efficiency and, in turn, job satisfaction.
ContributorsClark, Carina M (Author) / Bernardi, Jose (Thesis advisor) / Heywood, William (Committee member) / Watts, Richard (Committee member) / Rosso, Rachel (Committee member) / Arizona State University (Publisher)
Created2013
Description
Japanese literature of the Heian Era (794-1185) abounds with references to musical instruments and episodes of performance. This thesis provides some insight into that music by translating sections of the "Wakana II" (Spring Shoots II) chapter of the early 11th-century novel Genji monogatari (The Tale of Genji). It explains the musical references and shows how, in the context of the novel, musical performance, musical teaching, and interpersonal relationships were inextricably intertwined. Detailed appendices provide background on traditional Japanese musical instruments, musical theory, and related subjects.
ContributorsBotway, Lloyd (Author) / Creamer, John (Thesis advisor) / Chambers, Anthony (Committee member) / Solis, Theodore (Committee member) / Arizona State University (Publisher)
Created2013
Description
With the increase in computing power and availability of data, there has never been a greater need to understand data and make decisions from it. Traditional statistical techniques may not be adequate to handle the size of today's data or the complexities of the information hidden within the data. Thus knowledge discovery by machine learning techniques is necessary if we want to better understand information from data. In this dissertation, we explore the topics of asymmetric loss and asymmetric data in machine learning and propose new algorithms as solutions to some of the problems in these topics. We also studied variable selection of matched data sets and proposed a solution when there is non-linearity in the matched data. The research is divided into three parts. The first part addresses the problem of asymmetric loss. A proposed asymmetric support vector machine (aSVM) is used to predict specific classes with high accuracy. aSVM was shown to produce higher precision than a regular SVM. The second part addresses asymmetric data sets where variables are only predictive for a subset of the predictor classes. Asymmetric Random Forest (ARF) was proposed to detect these kinds of variables. The third part explores variable selection for matched data sets. Matched Random Forest (MRF) was proposed to find variables that are able to distinguish case and control without the restrictions that exists in linear models. MRF detects variables that are able to distinguish case and control even in the presence of interaction and qualitative variables.
ContributorsKoh, Derek (Author) / Runger, George C. (Thesis advisor) / Wu, Tong (Committee member) / Pan, Rong (Committee member) / Cesta, John (Committee member) / Arizona State University (Publisher)
Created2013