Matching Items (233)
Filtering by
- All Subjects: Machine Learning
- Member of: Barrett, The Honors College Thesis/Creative Project Collection
- Resource Type: Text
- Status: Published
Leveraging Machine Learning and Wireless Sensing for Robot Localization - Location Variance Analysis
Description
Object localization is used to determine the location of a device, an important aspect of applications ranging from autonomous driving to augmented reality. Commonly-used localization techniques include global positioning systems (GPS), simultaneous localization and mapping (SLAM), and positional tracking, but all of these methodologies have drawbacks, especially in high traffic indoor or urban environments. Using recent improvements in the field of machine learning, this project proposes a new method of localization using networks with several wireless transceivers and implemented without heavy computational loads or high costs. This project aims to build a proof-of-concept prototype and demonstrate that the proposed technique is feasible and accurate.
Modern communication networks heavily depend upon an estimate of the communication channel, which represents the distortions that a transmitted signal takes as it moves towards a receiver. A channel can become quite complicated due to signal reflections, delays, and other undesirable effects and, as a result, varies significantly with each different location. This localization system seeks to take advantage of this distinctness by feeding channel information into a machine learning algorithm, which will be trained to associate channels with their respective locations. A device in need of localization would then only need to calculate a channel estimate and pose it to this algorithm to obtain its location.
As an additional step, the effect of location noise is investigated in this report. Once the localization system described above demonstrates promising results, the team demonstrates that the system is robust to noise on its location labels. In doing so, the team demonstrates that this system could be implemented in a continued learning environment, in which some user agents report their estimated (noisy) location over a wireless communication network, such that the model can be implemented in an environment without extensive data collection prior to release.
Modern communication networks heavily depend upon an estimate of the communication channel, which represents the distortions that a transmitted signal takes as it moves towards a receiver. A channel can become quite complicated due to signal reflections, delays, and other undesirable effects and, as a result, varies significantly with each different location. This localization system seeks to take advantage of this distinctness by feeding channel information into a machine learning algorithm, which will be trained to associate channels with their respective locations. A device in need of localization would then only need to calculate a channel estimate and pose it to this algorithm to obtain its location.
As an additional step, the effect of location noise is investigated in this report. Once the localization system described above demonstrates promising results, the team demonstrates that the system is robust to noise on its location labels. In doing so, the team demonstrates that this system could be implemented in a continued learning environment, in which some user agents report their estimated (noisy) location over a wireless communication network, such that the model can be implemented in an environment without extensive data collection prior to release.
ContributorsChang, Roger (Co-author) / Kann, Trevor (Co-author) / Alkhateeb, Ahmed (Thesis director) / Bliss, Daniel (Committee member) / Electrical Engineering Program (Contributor) / Barrett, The Honors College (Contributor)
Created2020-05
Description
At present, the vast majority of human subjects with neurological disease are still diagnosed through in-person assessments and qualitative analysis of patient data. In this paper, we propose to use Topological Data Analysis (TDA) together with machine learning tools to automate the process of Parkinson’s disease classification and severity assessment. An automated, stable, and accurate method to evaluate Parkinson’s would be significant in streamlining diagnoses of patients and providing families more time for corrective measures. We propose a methodology which incorporates TDA into analyzing Parkinson’s disease postural shifts data through the representation of persistence images. Studying the topology of a system has proven to be invariant to small changes in data and has been shown to perform well in discrimination tasks. The contributions of the paper are twofold. We propose a method to 1) classify healthy patients from those afflicted by disease and 2) diagnose the severity of disease. We explore the use of the proposed method in an application involving a Parkinson’s disease dataset comprised of healthy-elderly, healthy-young and Parkinson’s disease patients.
ContributorsRahman, Farhan Nadir (Co-author) / Nawar, Afra (Co-author) / Turaga, Pavan (Thesis director) / Krishnamurthi, Narayanan (Committee member) / Electrical Engineering Program (Contributor) / Computer Science and Engineering Program (Contributor) / Barrett, The Honors College (Contributor)
Created2020-05
Description
In this project, the use of deep neural networks for the process of selecting actions to execute within an environment to achieve a goal is explored. Scenarios like this are common in crafting based games such as Terraria or Minecraft. Goals in these environments have recursive sub-goal dependencies which form a dependency tree. An agent operating within these environments have access to low amounts of data about the environment before interacting with it, so it is crucial that this agent is able to effectively utilize a tree of dependencies and its environmental surroundings to make judgements about which sub-goals are most efficient to pursue at any point in time. A successful agent aims to minimizes cost when completing a given goal. A deep neural network in combination with Q-learning techniques was employed to act as the agent in this environment. This agent consistently performed better than agents using alternate models (models that used dependency tree heuristics or human-like approaches to make sub-goal oriented choices), with an average performance advantage of 33.86% (with a standard deviation of 14.69%) over the best alternate agent. This shows that machine learning techniques can be consistently employed to make goal-oriented choices within an environment with recursive sub-goal dependencies and low amounts of pre-known information.
ContributorsKoleber, Derek (Author) / Acuna, Ruben (Thesis director) / Bansal, Ajay (Committee member) / W.P. Carey School of Business (Contributor) / Software Engineering (Contributor) / Barrett, The Honors College (Contributor)
Created2018-05
Description
This thesis dives into the world of artificial intelligence by exploring the functionality of a single layer artificial neural network through a simple housing price classification example while simultaneously considering its impact from a data management perspective on both the software and hardware level. To begin this study, the universally accepted model of an artificial neuron is broken down into its key components and then analyzed for functionality by relating back to its biological counterpart. The role of a neuron is then described in the context of a neural network, with equal emphasis placed on how it individually undergoes training and then for an entire network. Using the technique of supervised learning, the neural network is trained with three main factors for housing price classification, including its total number of rooms, bathrooms, and square footage. Once trained with most of the generated data set, it is tested for accuracy by introducing the remainder of the data-set and observing how closely its computed output for each set of inputs compares to the target value. From a programming perspective, the artificial neuron is implemented in C so that it would be more closely tied to the operating system and therefore make the collected profiler data more precise during the program's execution. The program is designed to break down each stage of the neuron's training process into distinct functions. In addition to utilizing more functional code, the struct data type is used as the underlying data structure for this project to not only represent the neuron but for implementing the neuron's training and test data. Once fully trained, the neuron's test results are then graphed to visually depict how well the neuron learned from its sample training set. Finally, the profiler data is analyzed to describe how the program operated from a data management perspective on the software and hardware level.
ContributorsRichards, Nicholas Giovanni (Author) / Miller, Phillip (Thesis director) / Meuth, Ryan (Committee member) / Computer Science and Engineering Program (Contributor) / Barrett, The Honors College (Contributor)
Created2018-05
Description
The passage of 2007's Legal Arizona Workers Act, which required all new hires to be tested for legal employment status through the federal E-Verify database, drastically changed the employment prospects for undocumented workers in the state. Using data from the 2007-2010 American Community Survey, this paper seeks to identify the impact of this law on the labor force in Arizona, specifically regarding undocumented workers and less educated native workers. Overall, the data shows that the wage bias against undocumented immigrants doubled in the four years studied, and the wages of native workers without a high school degree saw a temporary, positive increase compared to comparable workers in other states. The law did not have an effect on the wages of native workers with a high school degree.
ContributorsSantiago, Maria Christina (Author) / Pereira, Claudiney (Thesis director) / Mendez, Jose (Committee member) / School of International Letters and Cultures (Contributor) / Department of Economics (Contributor) / Barrett, The Honors College (Contributor)
Created2018-05
Description
Sex trafficking is an issue that is prevalent in the United States, including in Arizona. The Catholic Charities Diversion Program in Phoenix seeks to rehabilitate adults who have been involved in prostitution. The aim of this paper was to test three pilot interventions to address stress experienced by the program clients through three different techniques that were given in workshop format. Cognitive Behavior Therapy (CBT), Dance Movement Therapy, and yoga and meditation are the three types of stress reduction techniques that have been studied in previous research and were pilot tested with adult sex trafficking victims. The results of the pilot studies and all three techniques reduced levels of stress significantly, and they warrant future testing.
ContributorsSatapathy, Nikita (Co-author) / Khanal, Garima (Co-author) / Somayaji, Vallari (Co-author) / Roe-Sepowitz, Dominique (Thesis director) / Graff, Sarah (Committee member) / School of Life Sciences (Contributor) / Barrett, The Honors College (Contributor)
Created2016-05
Description
Almost every form of cancer deregulates the expression and activity of anabolic glycosyltransferase (GT) enzymes, which incorporate particular monosaccharides in a donor acceptor as well as linkage- and anomer-specific manner to assemble complex and diverse glycans that significantly affect numerous cellular events, including tumorigenesis and metastasis. Because glycosylation is not template-driven, GT deregulation yields heterogeneous arrays of aberrant intact glycan products, some in undetectable quantities in clinical bio-fluids (e.g., blood plasma). Numerous glycan features (e.g., 6 sialylation, β-1,6-branching, and core fucosylation) stem from approximately 25 glycan “nodes:” unique linkage specific monosaccharides at particular glycan branch points that collectively confer distinguishing features upon glycan products. For each node, changes in normalized abundance (Figure 1) may serve as nearly 1:1 surrogate measure of activity for culpable GTs and may correlate with particular stages of carcinogenesis. Complementary to traditional top down glycomics, the novel bottom-up technique applied herein condenses each glycan node and feature into a single analytical signal, quantified by two GC-MS instruments: GCT (time-of-flight analyzer) and GCMSD (transmission quadrupole analyzers). Bottom-up analysis of stage 3 and 4 breast cancer cases revealed better overall precision for GCMSD yet comparable clinical performance of both GC MS instruments and identified two downregulated glycan nodes as excellent breast cancer biomarker candidates: t-Gal and 4,6-GlcNAc (ROC AUC ≈ 0.80, p < 0.05). Resulting from the activity of multiple GTs, t-Gal had the highest ROC AUC (0.88) and lowest ROC p‑value (0.001) among all analyzed nodes. Representing core-fucosylation, glycan node 4,6-GlcNAc is a nearly 1:1 molecular surrogate for the activity of α-(1,6)-fucosyltransferase—a potential target for cancer therapy. To validate these results, future projects can analyze larger sample sets, find correlations between breast cancer stage and changes in t-Gal and 4,6-GlcNAc levels, gauge the specificity of these nodes for breast cancer and their potential role in other cancer types, and develop clinical tests for reliable breast cancer diagnosis and treatment monitoring based on t-Gal and 4,6-GlcNAc.
ContributorsZaare, Sahba (Author) / Borges, Chad (Thesis director) / LaBaer, Joshua (Committee member) / School of Molecular Sciences (Contributor) / Barrett, The Honors College (Contributor)
Created2016-05
Description
Social impact bonds (SIBs) are a multi-year contract between social service providers, the government, and private investors. The three parties agree on a specific outcome for a societal issue. Investors provide capital required for the service provider to operate the project. The service provider then delivers the service to the target population. The success of the project is evaluated by outside party. If the target outcome is met, the government repays the investors at a premium. Nonprofit service providers can only serve a small community as they lack the funding to scale their programs and their reliance on government funding and philanthropy leads to a lot of time focused on raising money in the short-term and inhibits them from evolving their programs and projects for long-term strategic success. Government budgets decline but social problems persist. These contracts share risk between the government and the investors and allow governments to test out programs and alleviate taxpayer burdens from unsuccessful social service programs. Arizona has a severe homelessness problem. Nightly, 6000 people are homeless in Maricopa County. In a given year, over 32,000 individuals were homeless, composed of single adults, families, children, and veterans. Homelessness is not only a debilitating and difficult experience for those who experience it, but also has considerable economic costs on society. Homeless individuals use a number of government programs beyond emergency shelters, and these can cost taxpayers billions of dollars per year. Rapid rehousing was a successful intervention model that the state has been heavily investing in the last few years. This thesis aimed to survey the Arizona climate and determine what barriers were present for enacting an SIB for homelessness. The findings showed that although there are many competent stakeholder groups, lack of interest and overall knowledge of SIBs prevented groups from taking responsibility as the anchor for such a project. Additionally, the government and nonprofits had good partnerships, but lacked relationships with the business community and investors that could propel an SIB. Finally, although rapid rehousing can be used as a successful intervention model, there are not enough years of proven success to justify the spending on an SIB. Additionally, data collection for homelessness programming needs to be standardized between all relevant partners. The framework for an SIB exists in Arizona, but needs a few more years of development before it can be considered.
ContributorsAhmed, Fabeeha (Author) / Desouza, Kevin (Thesis director) / Lucio, Joanna (Committee member) / School of Politics and Global Studies (Contributor) / Department of Economics (Contributor) / Barrett, The Honors College (Contributor)
Created2016-05
Description
In Arizona's early history, Females garnered more independence than most other women in the United States because they were forced to build a completely new life in settlements with little to no infrastructure. Now, Arizona has achieved a level of equality that no other state has yet to achieve in regard to gender representation. Yet, we have yet to achieve total equity. This paper looks to analyze responses that female senators from the Arizona State Legislature gave while being interviewed by the author. With questions derived from previous research conducted on women in politics at the state and federal level, this paper will delve into the personal experiences of six female senators. Although their personal narratives differ, their stories seem to reflect a collective tie that unites the female members together, beyond party allegiance. Each of the responses given by the senators had some aspects that showed trends supporting the majority of the hypotheses. Moving forward, in order to achieve 50% equality, two more senators would need to be elected and replace male senators.
ContributorsMacdonald, James Nicholson (Author) / Woodall, Gina (Thesis director) / Lyon, Jenna (Committee member) / School of Politics and Global Studies (Contributor) / School of Sustainability (Contributor) / Barrett, The Honors College (Contributor)
Created2016-05
Description
In 2012, President Obama presented an executive order, Deferred Action for Childhood Arrivals (DACA), which primarily defers the deportation of unauthorized immigrants who are under the age of 31 and who arrived to the US before the age 16, among other things. This study examines the impact DACA has had on the identity formation, interfamilial relationships, and future plans of 15 Mexican-origin adults (18-27 years old, 47% female) who were approved for DACA in Arizona. Participants were recruited using flyers and the snowball sampling method. Interviews were conducted using a semi-structured interview guide by a bilingual, culturally competent interviewer. The interviews were recorded and then a transcript-based pragmatic, thematic analysis of the interviews was conducted. Findings show that participants who arrive to the U.S. at a younger age (under 5) identify as American, while those who arrive at an older age (over 8) do not feel like they can identify as an American because they spent more time in Mexico and are more attached to their home culture. Physical characteristics also played a factor in whether or not participants felt like they could identify as American. Participants describe their financial responsibility in their families increasing since receiving DACA. They also describe how they are now seen as role models to other undocumented youth in their families. Despite the uncertain future of DACA, these participants continued to have ambitious goals such as becoming lawyers and working at robotics companies. Future studies should include larger sample sizes and formally test theories of identity.
ContributorsLopez, Alejandra (Author) / Diaz McConnell, Eileen (Thesis director) / Martinez, Airin (Committee member) / School of Social Transformation (Contributor) / School of Transborder Studies (Contributor) / Barrett, The Honors College (Contributor)
Created2016-05