Matching Items (157)
Filtering by
Description
Traditional Reinforcement Learning (RL) assumes to learn policies with respect to reward available from the environment but sometimes learning in a complex domain requires wisdom which comes from a wide range of experience. In behavior based robotics, it is observed that a complex behavior can be described by a combination of simpler behaviors. It is tempting to apply similar idea such that simpler behaviors can be combined in a meaningful way to tailor the complex combination. Such an approach would enable faster learning and modular design of behaviors. Complex behaviors can be combined with other behaviors to create even more advanced behaviors resulting in a rich set of possibilities. Similar to RL, combined behavior can keep evolving by interacting with the environment. The requirement of this method is to specify a reasonable set of simple behaviors. In this research, I present an algorithm that aims at combining behavior such that the resulting behavior has characteristics of each individual behavior. This approach has been inspired by behavior based robotics, such as the subsumption architecture and motor schema-based design. The combination algorithm outputs n weights to combine behaviors linearly. The weights are state dependent and change dynamically at every step in an episode. This idea is tested on discrete and continuous environments like OpenAI’s “Lunar Lander” and “Biped Walker”. Results are compared with related domains like Multi-objective RL, Hierarchical RL, Transfer learning, and basic RL. It is observed that the combination of behaviors is a novel way of learning which helps the agent achieve required characteristics. A combination is learned for a given state and so the agent is able to learn faster in an efficient manner compared to other similar approaches. Agent beautifully demonstrates characteristics of multiple behaviors which helps the agent to learn and adapt to the environment. Future directions are also suggested as possible extensions to this research.
ContributorsVora, Kevin Jatin (Author) / Zhang, Yu (Thesis advisor) / Yang, Yezhou (Committee member) / Praharaj, Sarbeswar (Committee member) / Arizona State University (Publisher)
Created2021
Description
The focus of my honors thesis is to find ways to use deep learning in tandem with tools in statistical mechanics to derive new ways to solve problems in biophysics. More specifically, I’ve been interested in finding transition pathways between two known states of a biomolecule. This is because understanding the mechanisms in which proteins fold and ligands bind is crucial to creating new medicines and understanding biological processes. In this thesis, I work with individuals in the Singharoy lab to develop a formulation to utilize reinforcement learning and sampling-based robotics planning to derive low free energy transition pathways between two known states. Our formulation uses Jarzynski’s equality and the stiff-spring approximation to obtain point estimates of energy, and construct an informed path search with atomistic resolution. At the core of this framework, is our first ever attempt we use a policy driven adaptive steered molecular dynamics (SMD) to control our molecular dynamics simulations. We show that both the reinforcement learning (RL) and robotics planning realization of the RL-guided framework can solve for pathways on toy analytical surfaces and alanine dipeptide.
ContributorsHo, Nicholas (Author) / Maciejewski, Ross (Thesis director) / Singharoy, Abhishek (Committee member) / Barrett, The Honors College (Contributor) / School of Mathematical and Statistical Sciences (Contributor) / Computer Science and Engineering Program (Contributor)
Created2022-12
Description
Molecular Dynamics (MD) simulations are ubiquitous throughout the physical sci-ences; they are critical in understanding how particle structures evolve over time
given a particular energy function. A software package called ParSplice introduced a
new method to generate these simulations in parallel that has significantly inflated
their length. Typically, simulations are short discrete Markov chains, only captur-
ing a few microseconds of a particle’s behavior and containing tens of thousands of
transitions between states; in contrast, a typical ParSplice simulation can be as long
as a few milliseconds, containing tens of millions of transitions. Naturally, sifting
through data of this size is impossible by hand, and there are a number of visualiza-
tion systems that provide comprehensive and intuitive analyses of particle structures
throughout MD simulations. However, no visual analytics systems have been built
that can manage the simulations that ParSplice produces. To analyze these large
data-sets, I built a visual analytics system that provides multiple coordinated views
that simultaneously describe the data temporally, within its structural context, and
based on its properties. The system provides fluid and powerful user interactions
regardless of the size of the data, allowing the user to drill down into the data-set to
get detailed insights, as well as run and save various calculations, most notably the
Nudged Elastic Band method. The system also allows the comparison of multiple
trajectories, revealing more information about the general behavior of particles at different temperatures, energy states etc.
ContributorsHnatyshyn, Rostyslav (Author) / Maciejewski, Ross (Thesis advisor) / Bryan, Chris (Committee member) / Ahrens, James (Committee member) / Arizona State University (Publisher)
Created2022
Description
Machine learning models are increasingly being deployed in real-world applications where their predictions are used to make critical decisions in a variety of domains. The proliferation of such models has led to a burgeoning need to ensure the reliability and safety of these models, given the potential negative consequences of model vulnerabilities. The complexity of machine learning models, along with the extensive data sets they analyze, can result in unpredictable and unintended outcomes. Model vulnerabilities may manifest due to errors in data input, algorithm design, or model deployment, which can have significant implications for both individuals and society. To prevent such negative outcomes, it is imperative to identify model vulnerabilities at an early stage in the development process. This will aid in guaranteeing the integrity, dependability, and safety of the models, thus mitigating potential risks and enabling the full potential of these technologies to be realized. However, enumerating vulnerabilities can be challenging due to the complexity of the real-world environment. Visual analytics, situated at the intersection of human-computer interaction, computer graphics, and artificial intelligence, offers a promising approach for achieving high interpretability of complex black-box models, thus reducing the cost of obtaining insights into potential vulnerabilities of models. This research is devoted to designing novel visual analytics methods to support the identification and analysis of model vulnerabilities. Specifically, generalizable visual analytics frameworks are instantiated to explore vulnerabilities in machine learning models concerning security (adversarial attacks and data perturbation) and fairness (algorithmic bias). In the end, a visual analytics approach is proposed to enable domain experts to explain and diagnose the model improvement of addressing identified vulnerabilities of machine learning models in a human-in-the-loop fashion. The proposed methods hold the potential to enhance the security and fairness of machine learning models deployed in critical real-world applications.
ContributorsXie, Tiankai (Author) / Maciejewski, Ross (Thesis advisor) / Liu, Huan (Committee member) / Bryan, Chris (Committee member) / Tong, Hanghang (Committee member) / Arizona State University (Publisher)
Created2023
Description
Although Saudi Arabia is moving towards a sustainable future, Existing residential buildings in the country are extremely unsustainable. Therefore, there is a necessity for greening the existing residential building. Mostadam green rating systems was developed by the Saudi ministry of housing in 2019 to address the long-term sustainability vision in residential buildings in the country. By setting Mostadam requirements as an objective of the retrofit process, it will ensure that the building achieve sustainability. However, Mostadam is new and there is a lack of knowledge of implementing its requirements on existing buildings. The aim of this research is to develop a framework to green existing residential buildings in Saudi Arabia to achieve Mostadam energy and water minimum requirements. The framework was developed based on an extensive keyword-based search and an analysis of 92 relevant research. The process starts with assessing the building against the minimum requirements of energy and water of Mostadam. After that, optimization phase is conducted. Building information modelling is used in the optimization phase. Energy and water efficiency optimization measures are identified from the analysed literature. Revit is used in the base model authoring and Green building studio cloud is used to simulate the energy and water efficiency measures. Then, payback period is calculated for all the efficiency measured to assess the decision making. A case study of a villa in Riyadh, Saudi Arabia is provided. result shows that the implemented efficiency measures led to an increment of 37.5% in annual energy savings and 26.1% in the annual water savings. Results shows that the application of the proposed framework supports evaluating energy and water efficiency measures to implement it on the buildings to achieve Mostadam minimum energy and water requirements. Recommendations were made for future work to bridge the knowledge gap.
ContributorsMohamed, Sara Murad (Author) / Sullivan, Kenneth (Thesis advisor) / Chong, Oswald (Committee member) / Hurtado, Kristen (Committee member) / Arizona State University (Publisher)
Created2022
Description
As the construction industry in Saudi Arabia was on its way to thriving again. Their growth was due to the unprecedented volume of planned projects such as large-scale and unique projects. Suddenly, the world was faced with one of the most disrupting events in the last century which had a devastating impact on the construction industry specifically. This paper explores mainly the impact of the COVID-19 pandemic on construction projects in Saudi Arabia. Particularly, this paper explores how the pandemic and its related events contributed to the projects' schedule disturbances. This is because most of the projects rely on manpower and supply chains which were heavily disrupted due to the protective measures. For that, a study was conducted to evaluate the impact on the construction projects in Saudi Arabia, to what extent the schedule projects were affected, and what were the main reasons for the schedule delays. The research relied on a field survey and schedule analysis for 12 projects which resulted in identifying several causes of delays and the delayed durations that the projects in Saudi Arabia were facing. This research allows those in construction fields to identify the main causes of delays in order to avoid or minimize the impact of these issues on future projects.
ContributorsObeid, Muhammad Hasan Hani (Author) / Ariaratnam, Samuel (Thesis advisor) / El Asmar, Mounir (Committee member) / Chong, Oswald (Committee member) / Arizona State University (Publisher)
Created2021
Description
Cities in the Global South face rapid urbanization challenges and often suffer an acute lack of infrastructure and governance capacities. Smart Cities Mission, in India, launched in 2015, aims to offer a novel approach for urban renewal of 100 cities following an area‐based development approach, where the use of ICT and digital technologies is particularly emphasized. This article presents a critical review of the design and implementation framework of this new urban renewal program across selected case‐study cities. The article examines the claims of the so‐called “smart cities” against actual urban transformation on‐ground and evaluates how “inclusive” and “sustainable” these developments are. We quantify the scale and coverage of the smart city urban renewal projects in the cities to highlight who the program includes and excludes. The article also presents a statistical analysis of the sectoral focus and budgetary allocations of the projects under the Smart Cities Mission to find an inherent bias in these smart city initiatives in terms of which types of development they promote and the ones it ignores. The findings indicate that a predominant emphasis on digital urban renewal of selected precincts and enclaves, branded as “smart cities,” leads to deepening social polarization and gentrification. The article offers crucial urban planning lessons for designing ICT‐driven urban renewal projects, while addressing critical questions around inclusion and sustainability in smart city ventures.`
ContributorsPraharaj, Sarbeswar (Author)
Created2021-05-07
Description
The consequences of failures from large-diameter water pipelines can be severe. Results can include significant property damage, damage to adjacent infrastructure such as roads and bridges resulting in transportation delays or shutdowns, adjacent structural damage to buildings resulting in loss of business, service disruption to a significant number of customers, loss of water, costly emergency repairs, and even loss of life. The American Water Works Association’s (AWWA) 2020 “State of the Water Industry” report states the top issue facing the water industry since 2016 is aging infrastructure, with the second being financing for improvements. The industry must find innovative ways to extend asset life and reduce maintenance expenditures. While are many different assets comprise the drinking water industry, pipelines are a major component and often neglected because they are typically buried. Reliability Centered Maintenance (RCM) is a process used to determine the most effective maintenance strategy for an asset, with the ultimate goal being to establish the required function of the asset with the required reliability at the lowest operations and maintenance costs. The RCM philosophy considers Preventive Maintenance, Predictive Maintenance, Condition Based Monitoring, Reactive Maintenance, and Proactive Maintenance techniques in an integrated manner to increase the probability an asset will perform its designed function throughout its design life with minimal maintenance. In addition to determining maintenance tasks, the timely performance of those tasks is crucial. If performed too late an asset may fail; if performed too early, resources that may be used better elsewhere are expended. Utility agencies can save time and money by using RCM analysis for their drinking water infrastructure. This dissertation reviews industries using RCM, discusses the benefits of an RCM analysis, and goes through a case study of an RCM at a large aqueduct in the United States. The dissertation further discusses the consequence of failure of large diameter water pipelines and proposes a regression model to help agencies determine the optimum time to perform maintenance tasks on large diameter prestressed concrete pipelines using RCM analysis.
ContributorsGeisbush, James R (Author) / Ariaratnam, Samuel T (Thesis advisor) / Grau, David (Committee member) / Chong, Oswald (Committee member) / Arizona State University (Publisher)
Created2024
Description
The world faces significant environmental and social challenges due to high economic development, population growth, industrialization, rapid urbanization, and unsustainable consumption. Global communities are taking the necessary measures to confront these international challenges and applying sustainable development principles across all sectors. Construction is a critical driving instrument of economic activity, and to achieve sustainable development, it is vital to transform conventional construction into a more sustainable model. The research investigated sustainable construction perceptions in Kuwait, a rapidly growing country with a high volume of construction activities. Kuwait has ambitious plans to transition into a more sustainable economic development model, and the construction industry needs to align with these plans. This research aims to identify the characteristics of sustainable construction applications in the Kuwaiti construction market, such as awareness, current perceptions, drivers and barriers, and the construction regulations' impact. The research utilized a qualitative approach to answer research questions and deliver research objectives by conducting eleven Semi-structured interviews with experienced professionals in the Kuwaiti construction market to collect rich data that reflects insights and understandings of the Kuwaiti construction industry. The Thematic analysis of the data resulted in six themes and one sub-theme that presented reflections, insights, and perspectives on sustainable construction perceptions in the Kuwaiti construction market. The research findings reflected poor sustainable construction awareness and poor environmental and social application in the construction industry, the determinant role of construction regulations in promoting sustainable construction. and barriers and drivers to sustainable construction applications. The research concluded with answers to research questions, delivery of research objectives, and an explanation of sustainable construction perceptions in the Kuwaiti construction market.
Contributorsalsalem, mohammad salem (Author) / Duran, Melanie (Thesis advisor) / Chong, Oswald (Committee member) / Sullivan, Kenneth (Committee member) / Grau, David (Committee member) / Arizona State University (Publisher)
Created2023
Description
Artificial Intelligence (AI) systems have achieved outstanding performance and have been found to be better than humans at various tasks, such as sentiment analysis, and face recognition. However, the majority of these state-of-the-art AI systems use complex Deep Learning (DL) methods which present challenges for human experts to design and evaluate such models with respect to privacy, fairness, and robustness. Recent examination of DL models reveals that representations may include information that could lead to privacy violations, unfairness, and robustness issues. This results in AI systems that are potentially untrustworthy from a socio-technical standpoint. Trustworthiness in AI is defined by a set of model properties such as non-discriminatory bias, protection of users’ sensitive attributes, and lawful decision-making. The characteristics of trustworthy AI can be grouped into three categories: Reliability, Resiliency, and Responsibility. Past research has shown that the successful integration of an AI model depends on its trustworthiness. Thus it is crucial for organizations and researchers to build trustworthy AI systems to facilitate the seamless integration and adoption of intelligent technologies. The main issue with existing AI systems is that they are primarily trained to improve technical measures such as accuracy on a specific task but are not considerate of socio-technical measures. The aim of this dissertation is to propose methods for improving the trustworthiness of AI systems through representation learning. DL models’ representations contain information about a given input and can be used for tasks such as detecting fake news on social media or predicting the sentiment of a review. The findings of this dissertation significantly expand the scope of trustworthy AI research and establish a new paradigm for modifying data representations to balance between properties of trustworthy AI. Specifically, this research investigates multiple techniques such as reinforcement learning for understanding trustworthiness in users’ privacy, fairness, and robustness in classification tasks like cyberbullying detection and fake news detection. Since most social measures in trustworthy AI cannot be used to fine-tune or train an AI model directly, the main contribution of this dissertation lies in using reinforcement learning to alter an AI system’s behavior based on non-differentiable social measures.
ContributorsMosallanezhad, Ahmadreza (Author) / Liu, Huan (Thesis advisor) / Mancenido, Michelle (Thesis advisor) / Doupe, Adam (Committee member) / Maciejewski, Ross (Committee member) / Arizona State University (Publisher)
Created2023