Matching Items (118)
Filtering by
- Creators: Yang, Yezhou
- Member of: ASU Electronic Theses and Dissertations
Description
This study explores the ways in which LGBTQ young adults describe the aspects of their identities, and how those identities shape their service needs and experiences. A participatory action research component was explored as a research and service approach that is sensitive to LGBTQ young people living at the intersections of multiple identities. Although it is understood that LGBTQ young people come from a variety of backgrounds, research is limited in its understanding and exploration of how aspects of identity, such as race and class, influence the lives and service needs of this population. The data was collected through an initial set of interviews with fifteen LGBTQ-identified young adults ages 18 to 24. The interviewees were recruited from an LGBTQ youth-serving organization using a purposive sampling approach to reflect racial/ethnic and gender identity diversity. Following the interviews, eight of the participants engaged as co-researchers on a participatory action research (PAR) team for sixteen weeks. The process of this team's work was assessed through a reflective analysis to identify factors that impacted the participants' lives. Analysis of the interviews identified key themes related to identity among the LGBTQ young people. The interviewees experienced a multiplicity of identities that were both socially and individually constructed. These identities were impacted by their immediate and social environments. The young people also identified ways that they used their identities to influence their environments and enhance their own resilience. The service experiences and needs of the LGBTQ young people in this study were directly influenced by their multiple identities. Implications for intersectional approaches to serving this population are explored. Analysis of the PAR process identified four areas in which the young people were most impacted through their work and interactions with one another: relationships, communication, participation, and inclusion. Implications for research and service approaches with LGBTQ young people are discussed.
ContributorsWagaman, M. Alex (Author) / Segal, Elizabeth A. (Thesis advisor) / Adelman, Madelaine (Committee member) / Ayón, Cecilia (Committee member) / Arizona State University (Publisher)
Created2013
Description
Families with intellectually disabled caregivers are more likely than families without intellectually disabled caregivers to experience poor child welfare outcomes, including high rates of substantiation. However, little research has examined child maltreatment re-reports among this population. The objectives of this study were to begin to address this gap by examining maltreatment re-report rates, and factors associated with maltreatment re-reports, among child welfare-involved families with intellectually disabled caregivers. Survival analysis was conducted using restricted release data from the National Survey of Child and Adolescent Well-Being (NSCAW) to examine the hazard rate and survival rate of maltreatment re-reports for cases with, and without, intellectually disabled caregivers. Multivariate discrete-time hazard models were run using logistic regression to examine the relationship between various predictors and the hazard of maltreatment re-reports. Results revealed that child protection cases involving caregivers with intellectual disabilities were no more likely than cases without intellectually disabled caregivers to experience maltreatment re-reports. Predictors of maltreatment re-reports varied based on whether or not a case involved a caregiver with an intellectual disability. Child gender, child disability, and child race/ethnicity were significant predictors for cases involving caregivers with intellectual disabilities, whereas prior involvement with CPS, caretaker drug problems, and initial allegation substantiation were significant predictors for cases not involving caregivers with intellectual disabilities. These preliminary findings suggest that prevention, screening, and intervention strategies should consider variability of predictive factors based on caregiver intellectual disability status.
ContributorsJames, Stephen (Author) / Shafer, Michael S (Thesis advisor) / Krysik, Judy (Committee member) / Ayón, Cecilia (Committee member) / Arizona State University (Publisher)
Created2014
Description
Dating violence in ethnic minority populations is an understudied phenomenon and little attention has been paid to the experiences of Mexican American youth; less research has been done on how those experiences alter perceptions and acceptance of participation in prevention programs. This study advances knowledge on how Mexican American adolescents view dating violence prevention programs and how cultural beliefs and values may hinder or encourage effective participation. Focus groups (N = 9) were form with Mexican American youth aged 15-17 years separated by gender and acculturation status (Mexican Oriented/Bicultural/Anglo Oriented), as determined previously by acculturation scores measured by the Acculturation Rating Scale for Mexican Americans (ARMSA; 0 or below = Mexican Oriented, 0-1 = Bicultural, 1 or above = Anglo Oriented). Several themes emerged throughout the focus group discussions that were derived from culturally-based needs. Mexican American adolescents made recommendations for program development (e.g., a broad curriculum beyond the topic of dating violence) and delivery (e.g., barriers to participation, the implications of peer involvement) within the context of their cultural values and needs. Low acculturated and bicultural teens identified specific cultural needs and their relevance within a dating violence prevention program. However, across all groups, adolescents felt that the needs of Mexican American youth were similar to other youth in regards to dating violence prevention programs. Implications for how social work can best design and implement prevention programs for Mexican American adolescents are discussed.
ContributorsAltamirano, Bianca (Author) / Williams, Lela R (Thesis advisor) / Ayón, Cecilia (Committee member) / Marsiglia, Flavio F (Committee member) / Arizona State University (Publisher)
Created2011
Description
While the suicide rate in Mexico is relatively low when compared to countries throughout the world, it is increasing at an alarming pace. Unfortunately, the amount of suicide research focused on Mexican populations is relatively scarce. Using a sample of high school students living in Guanajuato, Mexico, this study explored the relationship between recent suicidal ideation and three factors that previous research in other countries has connected to suicide: Migration aspirations, religiosity, and sexual behavior. Using multiple and logistic regression, the results indicated the following: 1) Recent suicidal ideation predicted increased migration aspirations, 2) higher levels of external religiosity predicted lower odds of recent suicidal ideation, and 3) stronger parent-child relationships predicted lower odds of recent suicidal ideation. The findings are discussed in light of the Bronfenbrenner's ecological systems theory, Bogenschneider's risk/protection model, and Stark's religious commitment theory.
ContributorsHoffman, Steven (Author) / Marsiglia, Flavio F (Thesis advisor) / Ayón, Cecilia (Committee member) / García-Pérez, Maria Hilda (Committee member) / Arizona State University (Publisher)
Created2011
Description
The current study sought to reevaluate Cass' Theory of sexual identity formation in terms of lesbian identity development over the past twenty years and how media acts as mediation in lesbian identity development. Ten semi-structured interviews were conducted with only nine useable transcripts analyzed for this thesis. This study is an explanatory investigation into linear stage theory, specifically Cass' theory, as well as the impact of media as a mediator during lesbian identity development. This study had three objectives 1) to gain an understanding of the theory and its components related to lesbian identity development 2) to understand the lesbian identity formation process and 3) to understand the impact and influence if any, media has had on lesbian self-reported identity development. Qualitative methods were used to obtain information and analyze the responses. Results indicate that the participants in this study believed that the coming out process was important. This study's results showed that several of the participants entered each stage of the theory, while others did not. Media had little influence on the identity development, and the participants had mixed reviews of medias portrayal of lesbians. Implications for practice and further research are discussed.
ContributorsHaseley, Hilary (Author) / Lacasse, Jeffrey R (Thesis advisor) / Segal, Elizabeth (Committee member) / Rounds, Tamara (Committee member) / Arizona State University (Publisher)
Created2011
Description
ABSTRACT Research suggests that there are benefits of early intervention and in focusing on mental health for infants and toddlers who have been maltreated. Court Teams for Infants and Toddlers is a model program designed to improve developmental outcomes using a systemic change approach. Multi-system collaboration between the courts, child welfare, health professionals, child advocates, and community partners are promoted to increase awareness and improve outcomes for infants and toddlers who have been removed from their parents. The Court Teams model in Arizona is known as Best for Babies. This study looks at implementation efforts of Best for Babies in two counties, Yavapai and Pima, and the unique perspectives of foster parents and attorneys representing the infants and toddlers while in the foster care system. It is important for purposes of effective program implementation to understand whether the Best for Babies program has impacted how these stakeholders address the unique needs of infants and toddlers. Findings reveal that most foster parents in this study were not familiar with the Best for Babies program; however, many of the comments shared are aligned with the values of the program. For example, all participants commented that collaboration among various stakeholders is necessary. Areas of opportunity were also illustrated in the findings regarding Best for Babies program implementation. For instance, the study found that even those foster parents familiar with the program could not attribute an impact on their care of infants and toddlers specifically to Best for Babies.
ContributorsWhite, Jennifer (Author) / Krysik, Judy (Thesis advisor) / Roe-Sepowitz, Dominique (Committee member) / Ayón, Cecilia (Committee member) / Arizona State University (Publisher)
Created2011
Description
In a collaborative environment where multiple robots and human beings are expected
to collaborate to perform a task, it becomes essential for a robot to be aware of multiple
agents working in its work environment. A robot must also learn to adapt to
different agents in the workspace and conduct its interaction based on the presence
of these agents. A theoretical framework was introduced which performs interaction
learning from demonstrations in a two-agent work environment, and it is called
Interaction Primitives.
This document is an in-depth description of the new state of the art Python
Framework for Interaction Primitives between two agents in a single as well as multiple
task work environment and extension of the original framework in a work environment
with multiple agents doing a single task. The original theory of Interaction
Primitives has been extended to create a framework which will capture correlation
between more than two agents while performing a single task. The new state of the
art Python framework is an intuitive, generic, easy to install and easy to use python
library which can be applied to use the Interaction Primitives framework in a work
environment. This library was tested in simulated environments and controlled laboratory
environment. The results and benchmarks of this library are available in the
related sections of this document.
to collaborate to perform a task, it becomes essential for a robot to be aware of multiple
agents working in its work environment. A robot must also learn to adapt to
different agents in the workspace and conduct its interaction based on the presence
of these agents. A theoretical framework was introduced which performs interaction
learning from demonstrations in a two-agent work environment, and it is called
Interaction Primitives.
This document is an in-depth description of the new state of the art Python
Framework for Interaction Primitives between two agents in a single as well as multiple
task work environment and extension of the original framework in a work environment
with multiple agents doing a single task. The original theory of Interaction
Primitives has been extended to create a framework which will capture correlation
between more than two agents while performing a single task. The new state of the
art Python framework is an intuitive, generic, easy to install and easy to use python
library which can be applied to use the Interaction Primitives framework in a work
environment. This library was tested in simulated environments and controlled laboratory
environment. The results and benchmarks of this library are available in the
related sections of this document.
ContributorsKumar, Ashish, M.S (Author) / Amor, Hani Ben (Thesis advisor) / Zhang, Yu (Committee member) / Yang, Yezhou (Committee member) / Arizona State University (Publisher)
Created2017
Description
Computer Vision as a eld has gone through signicant changes in the last decade.
The eld has seen tremendous success in designing learning systems with hand-crafted
features and in using representation learning to extract better features. In this dissertation
some novel approaches to representation learning and task learning are studied.
Multiple-instance learning which is generalization of supervised learning, is one
example of task learning that is discussed. In particular, a novel non-parametric k-
NN-based multiple-instance learning is proposed, which is shown to outperform other
existing approaches. This solution is applied to a diabetic retinopathy pathology
detection problem eectively.
In cases of representation learning, generality of neural features are investigated
rst. This investigation leads to some critical understanding and results in feature
generality among datasets. The possibility of learning from a mentor network instead
of from labels is then investigated. Distillation of dark knowledge is used to eciently
mentor a small network from a pre-trained large mentor network. These studies help
in understanding representation learning with smaller and compressed networks.
The eld has seen tremendous success in designing learning systems with hand-crafted
features and in using representation learning to extract better features. In this dissertation
some novel approaches to representation learning and task learning are studied.
Multiple-instance learning which is generalization of supervised learning, is one
example of task learning that is discussed. In particular, a novel non-parametric k-
NN-based multiple-instance learning is proposed, which is shown to outperform other
existing approaches. This solution is applied to a diabetic retinopathy pathology
detection problem eectively.
In cases of representation learning, generality of neural features are investigated
rst. This investigation leads to some critical understanding and results in feature
generality among datasets. The possibility of learning from a mentor network instead
of from labels is then investigated. Distillation of dark knowledge is used to eciently
mentor a small network from a pre-trained large mentor network. These studies help
in understanding representation learning with smaller and compressed networks.
ContributorsVenkatesan, Ragav (Author) / Li, Baoxin (Thesis advisor) / Turaga, Pavan (Committee member) / Yang, Yezhou (Committee member) / Davulcu, Hasan (Committee member) / Arizona State University (Publisher)
Created2017
Description
With the rise of the Big Data Era, an exponential amount of network data is being generated at an unprecedented rate across a wide-range of high impact micro and macro areas of research---from protein interaction to social networks. The critical challenge is translating this large scale network data into actionable information.
A key task in the data translation is the analysis of network connectivity via marked nodes---the primary focus of our research. We have developed a framework for analyzing network connectivity via marked nodes in large scale graphs, utilizing novel algorithms in three interrelated areas: (1) analysis of a single seed node via it’s ego-centric network (AttriPart algorithm); (2) pathway identification between two seed nodes (K-Simple Shortest Paths Multithreaded and Search Reduced (KSSPR) algorithm); and (3) tree detection, defining the interaction between three or more seed nodes (Shortest Path MST algorithm).
In an effort to address both fundamental and applied research issues, we have developed the LocalForcasting algorithm to explore how network connectivity analysis can be applied to local community evolution and recommender systems. The goal is to apply the LocalForecasting algorithm to various domains---e.g., friend suggestions in social networks or future collaboration in co-authorship networks. This algorithm utilizes link prediction in combination with the AttriPart algorithm to predict future connections in local graph partitions.
Results show that our proposed AttriPart algorithm finds up to 1.6x denser local partitions, while running approximately 43x faster than traditional local partitioning techniques (PageRank-Nibble). In addition, our LocalForecasting algorithm demonstrates a significant improvement in the number of nodes and edges correctly predicted over baseline methods. Furthermore, results for the KSSPR algorithm demonstrate a speed-up of up to 2.5x the standard k-simple shortest paths algorithm.
A key task in the data translation is the analysis of network connectivity via marked nodes---the primary focus of our research. We have developed a framework for analyzing network connectivity via marked nodes in large scale graphs, utilizing novel algorithms in three interrelated areas: (1) analysis of a single seed node via it’s ego-centric network (AttriPart algorithm); (2) pathway identification between two seed nodes (K-Simple Shortest Paths Multithreaded and Search Reduced (KSSPR) algorithm); and (3) tree detection, defining the interaction between three or more seed nodes (Shortest Path MST algorithm).
In an effort to address both fundamental and applied research issues, we have developed the LocalForcasting algorithm to explore how network connectivity analysis can be applied to local community evolution and recommender systems. The goal is to apply the LocalForecasting algorithm to various domains---e.g., friend suggestions in social networks or future collaboration in co-authorship networks. This algorithm utilizes link prediction in combination with the AttriPart algorithm to predict future connections in local graph partitions.
Results show that our proposed AttriPart algorithm finds up to 1.6x denser local partitions, while running approximately 43x faster than traditional local partitioning techniques (PageRank-Nibble). In addition, our LocalForecasting algorithm demonstrates a significant improvement in the number of nodes and edges correctly predicted over baseline methods. Furthermore, results for the KSSPR algorithm demonstrate a speed-up of up to 2.5x the standard k-simple shortest paths algorithm.
ContributorsFreitas, Scott (Author) / Tong, Hanghang (Thesis advisor) / Maciejewski, Ross (Committee member) / Yang, Yezhou (Committee member) / Arizona State University (Publisher)
Created2018
Description
The performance of most of the visual computing tasks depends on the quality of the features extracted from the raw data. Insightful feature representation increases the performance of many learning algorithms by exposing the underlying explanatory factors of the output for the unobserved input. A good representation should also handle anomalies in the data such as missing samples and noisy input caused by the undesired, external factors of variation. It should also reduce the data redundancy. Over the years, many feature extraction processes have been invented to produce good representations of raw images and videos.
The feature extraction processes can be categorized into three groups. The first group contains processes that are hand-crafted for a specific task. Hand-engineering features requires the knowledge of domain experts and manual labor. However, the feature extraction process is interpretable and explainable. Next group contains the latent-feature extraction processes. While the original feature lies in a high-dimensional space, the relevant factors for a task often lie on a lower dimensional manifold. The latent-feature extraction employs hidden variables to expose the underlying data properties that cannot be directly measured from the input. Latent features seek a specific structure such as sparsity or low-rank into the derived representation through sophisticated optimization techniques. The last category is that of deep features. These are obtained by passing raw input data with minimal pre-processing through a deep network. Its parameters are computed by iteratively minimizing a task-based loss.
In this dissertation, I present four pieces of work where I create and learn suitable data representations. The first task employs hand-crafted features to perform clinically-relevant retrieval of diabetic retinopathy images. The second task uses latent features to perform content-adaptive image enhancement. The third task ranks a pair of images based on their aestheticism. The goal of the last task is to capture localized image artifacts in small datasets with patch-level labels. For both these tasks, I propose novel deep architectures and show significant improvement over the previous state-of-art approaches. A suitable combination of feature representations augmented with an appropriate learning approach can increase performance for most visual computing tasks.
The feature extraction processes can be categorized into three groups. The first group contains processes that are hand-crafted for a specific task. Hand-engineering features requires the knowledge of domain experts and manual labor. However, the feature extraction process is interpretable and explainable. Next group contains the latent-feature extraction processes. While the original feature lies in a high-dimensional space, the relevant factors for a task often lie on a lower dimensional manifold. The latent-feature extraction employs hidden variables to expose the underlying data properties that cannot be directly measured from the input. Latent features seek a specific structure such as sparsity or low-rank into the derived representation through sophisticated optimization techniques. The last category is that of deep features. These are obtained by passing raw input data with minimal pre-processing through a deep network. Its parameters are computed by iteratively minimizing a task-based loss.
In this dissertation, I present four pieces of work where I create and learn suitable data representations. The first task employs hand-crafted features to perform clinically-relevant retrieval of diabetic retinopathy images. The second task uses latent features to perform content-adaptive image enhancement. The third task ranks a pair of images based on their aestheticism. The goal of the last task is to capture localized image artifacts in small datasets with patch-level labels. For both these tasks, I propose novel deep architectures and show significant improvement over the previous state-of-art approaches. A suitable combination of feature representations augmented with an appropriate learning approach can increase performance for most visual computing tasks.
ContributorsChandakkar, Parag Shridhar (Author) / Li, Baoxin (Thesis advisor) / Yang, Yezhou (Committee member) / Turaga, Pavan (Committee member) / Davulcu, Hasan (Committee member) / Arizona State University (Publisher)
Created2017