Matching Items (294)
Filtering by
- All Subjects: artificial intelligence
- Creators: Arizona State University
Description
The rapid escalation of technology and the widespread emergence of modern technological equipments have resulted in the generation of humongous amounts of digital data (in the form of images, videos and text). This has expanded the possibility of solving real world problems using computational learning frameworks. However, while gathering a large amount of data is cheap and easy, annotating them with class labels is an expensive process in terms of time, labor and human expertise. This has paved the way for research in the field of active learning. Such algorithms automatically select the salient and exemplar instances from large quantities of unlabeled data and are effective in reducing human labeling effort in inducing classification models. To utilize the possible presence of multiple labeling agents, there have been attempts towards a batch mode form of active learning, where a batch of data instances is selected simultaneously for manual annotation. This dissertation is aimed at the development of novel batch mode active learning algorithms to reduce manual effort in training classification models in real world multimedia pattern recognition applications. Four major contributions are proposed in this work: $(i)$ a framework for dynamic batch mode active learning, where the batch size and the specific data instances to be queried are selected adaptively through a single formulation, based on the complexity of the data stream in question, $(ii)$ a batch mode active learning strategy for fuzzy label classification problems, where there is an inherent imprecision and vagueness in the class label definitions, $(iii)$ batch mode active learning algorithms based on convex relaxations of an NP-hard integer quadratic programming (IQP) problem, with guaranteed bounds on the solution quality and $(iv)$ an active matrix completion algorithm and its application to solve several variants of the active learning problem (transductive active learning, multi-label active learning, active feature acquisition and active learning for regression). These contributions are validated on the face recognition and facial expression recognition problems (which are commonly encountered in real world applications like robotics, security and assistive technology for the blind and the visually impaired) and also on collaborative filtering applications like movie recommendation.
ContributorsChakraborty, Shayok (Author) / Panchanathan, Sethuraman (Thesis advisor) / Balasubramanian, Vineeth N. (Committee member) / Li, Baoxin (Committee member) / Mittelmann, Hans (Committee member) / Ye, Jieping (Committee member) / Arizona State University (Publisher)
Created2013
Description
The increasing popularity of Twitter renders improved trustworthiness and relevance assessment of tweets much more important for search. However, given the limitations on the size of tweets, it is hard to extract measures for ranking from the tweet's content alone. I propose a method of ranking tweets by generating a reputation score for each tweet that is based not just on content, but also additional information from the Twitter ecosystem that consists of users, tweets, and the web pages that tweets link to. This information is obtained by modeling the Twitter ecosystem as a three-layer graph. The reputation score is used to power two novel methods of ranking tweets by propagating the reputation over an agreement graph based on tweets' content similarity. Additionally, I show how the agreement graph helps counter tweet spam. An evaluation of my method on 16~million tweets from the TREC 2011 Microblog Dataset shows that it doubles the precision over baseline Twitter Search and achieves higher precision than current state of the art method. I present a detailed internal empirical evaluation of RAProp in comparison to several alternative approaches proposed by me, as well as external evaluation in comparison to the current state of the art method.
ContributorsRavikumar, Srijith (Author) / Kambhampati, Subbarao (Thesis advisor) / Davulcu, Hasan (Committee member) / Liu, Huan (Committee member) / Arizona State University (Publisher)
Created2013
Description
Nurses are using health information technology during patient care activities in acute care at an unprecedented rate. Previous literature has presented nurses' response to technology obstacles as a work-around, a negative behavior. Using a narrative inquiry in one hospital unit, this dissertation examines nurses' interactions when they encounter technology obstacles from a complexity science perspective. In this alternative view, outcomes are understood to emerge from tensions in the environment through nonlinear and self-organizing interactions. Innovation is a process of changing interaction patterns to bring about transformation in practices or products that have the potential to contribute to social wellbeing, such as better care. Innovation was found when nurses responded to health information technology obstacles with self-organizing interactions, sensitivity to initial conditions, multidirectionality, and their actions were influenced by a plethora of sets of rules. Nurses self-organized with co-workers to find a better way to deliver care to patients when using technology. Nurses rarely told others outside their work-group of the obstacles that occurred in their everyday interactions, including hospital-wide process improvement committees. Managers were infrequently consulted when nurses encountered technology obstacles, and often nurses did not find solutions to their obstacles when they contacted the Help Desk. Opportunities exist to facilitate interactions among nurses and other members of the organization to realize better use of health information technology that improves quality and safety while decreasing cost in the patient experience.
ContributorsLalley, Catherine (Author) / Malloch, Kathy (Thesis advisor) / Fleury, Julie (Committee member) / Danzig, Arnold (Committee member) / Arizona State University (Publisher)
Created2013
Description
ABSTRACT Caregiving studies generally do not focus on the post-caregiving phase of care, or African Americans post-caregivers (AAPCGs). This mixed-methods study guided by the Transitions Theory, explored the experiences of 40 AAPCGs residing in Los Angeles, California and Phoenix, Arizona, whose loved ones died within the last 10 years. Data collection tools included individual interviews, demographic questionnaire, CES-D, Brief Cope, and Social Support. Findings present the specific aims of the study. Aim 1 dealt with the types, patterns and properties of post-caregiving transitions (PCT). Many AAPCGs experienced multiple, simultaneous transitions that continued to impact their lives many years after caregiving ends. Aim 2 dealt with factors that facilitate or inhibit healthy PCT. Facilitators include: Being satisfied with care provided; fulfilling death-bed promises; living out the legacy of the deceased; deep spiritual beliefs in God and support of family, friends and church. Inhibitors include: Experiencing a deep sense of loss, confusion, depression, loneliness, and guilt; physical challenges such as fatigue and exhaustion, breathing problems, dizziness, fainting, cognitive difficulties, pain, headaches, hypertension and insomnia; family conflicts, job or home loss that linger long after PCT. Aim 3 involves process indicators including: connectedness with family, friends, co-workers, church and God; returning to work or school. Coping strategies that helped AAPCGs include: productive ventures, family mementoes, reminiscing, new baby, or visiting cemetery. Appropriate coping led to outcome indicators of mastery such as new environment; making decisions; taking actions; readying oneself for another caregiving role; preparing for one's own life and death; or caring for self. Fluid integrative identities include: Sense of balance, peacefulness and joy, fulfillment, compassion; remembering without pain; or new identity. Implications for practice, policy, education and research include: Care providers and policy makers must ensure that AA caregivers receive adequate EOL and hospice information and support for adequate preparation of loved one's death. Geriatric educators must design and implement curricular programming that includes the post-caregiving phase as a very important phase of caregiving. Researchers should design culturally-congruent assessment tools or improve the checklist developed in this study to appropriately measure PCT; and also develop culturally-relevant interventions to facilitate healthy PCT.
ContributorsUme, Ebere Peace (Author) / Evans, Bronwynne C. (Thesis advisor) / Coon, David W. (Thesis advisor) / Keller, Colleen S (Committee member) / Arizona State University (Publisher)
Created2013
Description
The healthcare system is plagued with increasing cost and poor quality outcomes. A major contributing factor for these issues is that outdated leadership practices, such as leader-centricity, linear thinking, and poor readiness for innovation, are being used in healthcare organizations. Through a qualitative case study analysis of innovation implementation, a new framework of leadership was uncovered. This framework presented new characteristics of leaders that led to the successful implementation of an innovation. Characteristics uncovered included boundary spanning, risk taking, visioning, leveraging opportunity, adaptation, coordination of information flow, and facilitation. These characteristics describe how leaders throughout the system were able to influence information flow, relationships, connections, and organizational context to implement innovation.
ContributorsWeberg, Daniel Robert (Author) / Fluery, Julie (Thesis advisor) / Malloch, Kathy (Thesis advisor) / Porter-O'Grady, Timothy (Committee member) / Hagler, Debra (Committee member) / Arizona State University (Publisher)
Created2013
Description
Objective of this thesis project is to build a prototype using Linear Temporal Logic specifications for generating a 2D motion plan commanding an iRobot to fulfill the specifications. This thesis project was created for Cyber Physical Systems Lab in Arizona State University. The end product of this thesis is creation of a software solution which can be used in the academia and industry for research in cyber physical systems related applications. The major features of the project are: creating a modular system for motion planning, use of Robot Operating System (ROS), use of triangulation for environment decomposition and using stargazer sensor for localization. The project is built on an open source software called ROS which provides an environment where it is very easy to integrate different modules be it software or hardware on a Linux based platform. Use of ROS implies the project or its modules can be adapted quickly for different applications as the need arises. The final software package created and tested takes a data file as its input which contains the LTL specifications, a symbols list used in the LTL and finally the environment polygon data containing real world coordinates for all polygons and also information on neighbors and parents of each polygon. The software package successfully ran the experiment of coverage, reachability with avoidance and sequencing.
ContributorsPandya, Parth (Author) / Fainekos, Georgios (Thesis advisor) / Dasgupta, Partha (Committee member) / Lee, Yann-Hang (Committee member) / Arizona State University (Publisher)
Created2013
Description
Linear Temporal Logic is gaining increasing popularity as a high level specification language for robot motion planning due to its expressive power and scalability of LTL control synthesis algorithms. This formalism, however, requires expert knowledge and makes it inaccessible to non-expert users. This thesis introduces a graphical specification environment to create high level motion plans to control robots in the field by converting a visual representation of the motion/task plan into a Linear Temporal Logic (LTL) specification. The visual interface is built on the Android tablet platform and provides functionality to create task plans through a set of well defined gestures and on screen controls. It uses the notion of waypoints to quickly and efficiently describe the motion plan and enables a variety of complex Linear Temporal Logic specifications to be described succinctly and intuitively by the user without the need for the knowledge and understanding of LTL specification. Thus, it opens avenues for its use by personnel in military, warehouse management, and search and rescue missions. This thesis describes the construction of LTL for various scenarios used for robot navigation using the visual interface developed and leverages the use of existing LTL based motion planners to carry out the task plan by a robot.
ContributorsSrinivas, Shashank (Author) / Fainekos, Georgios (Thesis advisor) / Baral, Chitta (Committee member) / Burleson, Winslow (Committee member) / Arizona State University (Publisher)
Created2013
Description
Answer Set Programming (ASP) is one of the most prominent and successful knowledge representation paradigms. The success of ASP is due to its expressive non-monotonic modeling language and its efficient computational methods originating from building propositional satisfiability solvers. The wide adoption of ASP has motivated several extensions to its modeling language in order to enhance expressivity, such as incorporating aggregates and interfaces with ontologies. Also, in order to overcome the grounding bottleneck of computation in ASP, there are increasing interests in integrating ASP with other computing paradigms, such as Constraint Programming (CP) and Satisfiability Modulo Theories (SMT). Due to the non-monotonic nature of the ASP semantics, such enhancements turned out to be non-trivial and the existing extensions are not fully satisfactory. We observe that one main reason for the difficulties rooted in the propositional semantics of ASP, which is limited in handling first-order constructs (such as aggregates and ontologies) and functions (such as constraint variables in CP and SMT) in natural ways. This dissertation presents a unifying view on these extensions by viewing them as instances of formulas with generalized quantifiers and intensional functions. We extend the first-order stable model semantics by by Ferraris, Lee, and Lifschitz to allow generalized quantifiers, which cover aggregate, DL-atoms, constraints and SMT theory atoms as special cases. Using this unifying framework, we study and relate different extensions of ASP. We also present a tight integration of ASP with SMT, based on which we enhance action language C+ to handle reasoning about continuous changes. Our framework yields a systematic approach to study and extend non-monotonic languages.
ContributorsMeng, Yunsong (Author) / Lee, Joohyung (Thesis advisor) / Ahn, Gail-Joon (Committee member) / Baral, Chitta (Committee member) / Fainekos, Georgios (Committee member) / Lifschitz, Vladimir (Committee member) / Arizona State University (Publisher)
Created2013
Description
Two studies were conducted to test a model to predict healthy lifestyle behaviors, physical activity, and body mass index (BMI) in Taiwanese adolescents by assessing their physical activity and nutrition knowledge, healthy lifestyle beliefs, and perceived difficulty in performing healthy lifestyle behaviors. The study drew upon cognitive behavioral theory to develop this study. The pilot study aimed to test and evaluate psychometric properties of eight Chinese-version scales. The total sample for the pilot study included 186 participants from two middle schools in Taiwan. The mean age was 13.19 for boys and 13.79 for girls. Most scales including Beck Youth Inventory self-concept, Beck Youth Inventory depression, Beck Youth Inventory anxiety, healthy lifestyle beliefs, perceived difficulty, and healthy lifestyle behaviors scales Cronbach alpha were above .90. The Cronbach alpha for the nutrition knowledge and the activity knowledge scale were .86 and .70, respectively. For the primary study, descriptive statistics were used to describe sample characteristics, and path analysis was used to test a model predicting BMI in Taiwanese adolescents. The total sample included 453 participants from two middle schools in Taiwan. The mean age of sample was 13.42 years; 47.5% (n = 215) were males. The mean BMI was 21.83 for boys and 19.84 for girls. The BMI for both boys and girls was within normal range. For path analysis, the chi-square was 426.82 (df = 22, p < .01). The CFI of .62 and the RMSEA of .20 suggested that the model had less than an adequate fit (Hu & Bentler, 1999). For alternative model, dropping the variable of gender from the model, the results indicated that it in fact was an adequate fit to the data (chi-square (23, 453) =33.75, p> .05; CFI= .98; RMSEA= .03). As expected, the results suggested that adolescents who reported higher healthy lifestyle beliefs had more healthy lifestyle behaviors. Furthermore, adolescents who perceived more difficulty in performing healthy lifestyle behaviors engaged in fewer healthy lifestyle behaviors and less physical activity. The findings suggested that adolescents' higher healthy lifestyle beliefs were positively associated with their healthy lifestyle behaviors.
ContributorsChan, Shu-Min (Author) / Melnyk, Bernadette Mazurek (Thesis advisor) / Belyea, Michael (Thesis advisor) / Chen, Angela Chia-Chen (Committee member) / Dodgson, Joan (Committee member) / Arizona State University (Publisher)
Created2012
Description
This thesis introduces the Model-Based Development of Multi-iRobot Toolbox (MBDMIRT), a Simulink-based toolbox designed to provide the means to acquire and practice the Model-Based Development (MBD) skills necessary to design real-time embedded system. The toolbox was developed in the Cyber-Physical System Laboratory at Arizona State University. The MBDMIRT toolbox runs under MATLAB/Simulink to simulate the movements of multiple iRobots and to control, after verification by simulation, multiple physical iRobots accordingly. It adopts the Simulink/Stateflow, which exemplifies an approach to MBD, to program the behaviors of the iRobots. The MBDMIRT toolbox reuses and augments the open-source MATLAB-Based Simulator for the iRobot Create from Cornell University to run the simulation. Regarding the mechanism of iRobot control, the MBDMIRT toolbox applies the MATLAB Toolbox for the iRobot Create (MTIC) from United States Naval Academy to command the physical iRobots. The MBDMIRT toolbox supports a timer in both the simulation and the control, which is based on the local clock of the PC running the toolbox. In addition to the build-in sensors of an iRobot, the toolbox can simulate four user-added sensors, which are overhead localization system (OLS), sonar sensors, a camera, and Light Detection And Ranging (LIDAR). While controlling a physical iRobot, the toolbox supports the StarGazer OLS manufactured by HAGISONIC, Inc.
ContributorsSu, Shih-Kai (Author) / Fainekos, Georgios E (Thesis advisor) / Sarjoughian, Hessam S. (Committee member) / Artemiadis, Panagiotis K (Committee member) / Arizona State University (Publisher)
Created2012