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 - 3 of 3
Filtering by
- Creators: Goryll, Michael
Description
Feature representations for raw data is one of the most important component in a machine learning system. Traditionally, features are \textit{hand crafted} by domain experts which can often be a time consuming process. Furthermore, they do not generalize well to unseen data and novel tasks. Recently, there have been many efforts to generate data-driven representations using clustering and sparse models. This dissertation focuses on building data-driven unsupervised models for analyzing raw data and developing efficient feature representations.
Simultaneous segmentation and feature extraction approaches for silicon-pores sensor data are considered. Aggregating data into a matrix and performing low rank and sparse matrix decompositions with additional smoothness constraints are proposed to solve this problem. Comparison of several variants of the approaches and results for signal de-noising and translocation/trapping event extraction are presented. Algorithms to improve transform-domain features for ion-channel time-series signals based on matrix completion are presented. The improved features achieve better performance in classification tasks and in reducing the false alarm rates when applied to analyte detection.
Developing representations for multimedia is an important and challenging problem with applications ranging from scene recognition, multi-media retrieval and personal life-logging systems to field robot navigation. In this dissertation, we present a new framework for feature extraction for challenging natural environment sounds. Proposed features outperform traditional spectral features on challenging environmental sound datasets. Several algorithms are proposed that perform supervised tasks such as recognition and tag annotation. Ensemble methods are proposed to improve the tag annotation process.
To facilitate the use of large datasets, fast implementations are developed for sparse coding, the key component in our algorithms. Several strategies to speed-up Orthogonal Matching Pursuit algorithm using CUDA kernel on a GPU are proposed. Implementations are also developed for a large scale image retrieval system. Image-based "exact search" and "visually similar search" using the image patch sparse codes are performed. Results demonstrate large speed-up over CPU implementations and good retrieval performance is also achieved.
Simultaneous segmentation and feature extraction approaches for silicon-pores sensor data are considered. Aggregating data into a matrix and performing low rank and sparse matrix decompositions with additional smoothness constraints are proposed to solve this problem. Comparison of several variants of the approaches and results for signal de-noising and translocation/trapping event extraction are presented. Algorithms to improve transform-domain features for ion-channel time-series signals based on matrix completion are presented. The improved features achieve better performance in classification tasks and in reducing the false alarm rates when applied to analyte detection.
Developing representations for multimedia is an important and challenging problem with applications ranging from scene recognition, multi-media retrieval and personal life-logging systems to field robot navigation. In this dissertation, we present a new framework for feature extraction for challenging natural environment sounds. Proposed features outperform traditional spectral features on challenging environmental sound datasets. Several algorithms are proposed that perform supervised tasks such as recognition and tag annotation. Ensemble methods are proposed to improve the tag annotation process.
To facilitate the use of large datasets, fast implementations are developed for sparse coding, the key component in our algorithms. Several strategies to speed-up Orthogonal Matching Pursuit algorithm using CUDA kernel on a GPU are proposed. Implementations are also developed for a large scale image retrieval system. Image-based "exact search" and "visually similar search" using the image patch sparse codes are performed. Results demonstrate large speed-up over CPU implementations and good retrieval performance is also achieved.
ContributorsSattigeri, Prasanna S (Author) / Spanias, Andreas (Thesis advisor) / Thornton, Trevor (Committee member) / Goryll, Michael (Committee member) / Tsakalis, Konstantinos (Committee member) / Arizona State University (Publisher)
Created2014
Description
Engineered nanoporous substrates made using materials such as silicon nitride or silica have been demonstrated to work as particle counters or as hosts for nano-lipid bilayer membrane formation. These mechanically fabricated porous structures have thicknesses of several hundred nanometers up to several micrometers to ensure mechanical stability of the membrane. However, it is desirable to have a three-dimensional structure to ensure increased mechanical stability. In this study, circular silica shells used from Coscinodiscus wailesii, a species of diatoms (unicellular marine algae) were immobilized on a silicon chip with a micrometer-sized aperture using a UV curable polyurethane adhesive. The current conducted by a single nanopore of 40 nm diameter and 50 nm length, during the translocation of a 27 nm polystyrene sphere was simulated using COMSOL multiphysics and tested experimentally. The current conducted by a single 40 nm diameter nanopore of the diatom shell during the translocation of a 27 nm polystyrene sphere was simulated using COMSOL Multiphysics (28.36 pA) and was compared to the experimental measurement (28.69 pA) and Coulter Counting theory (29.95 pA).In addition, a mobility of 1.11 x 10-8 m2s-1V-1 for the 27 nm polystyrene spheres was used to convert the simulated current from spatial dependence to time dependence.
To achieve a sensing diameter of 1-2 nanometers, the diatom shells were used as substrates to perform ion-channel reconstitution experiments. The immobilized diatom shell was functionalized using silane chemistry and lipid bilayer membranes were formed. Functionalization of the diatom shell surface improves bilayer formation probability from 1 out of 10 to 10 out of 10 as monitored by impedance spectroscopy. Self-insertion of outer membrane protein OmpF of E.Coli into the lipid membranes could be confirmed using single channel recordings, indicating that nano-BLMs had formed which allow for fully functional porin activity. The results indicate that biogenic silica nanoporous substrates can be simulated using a simplified two dimensional geometry to predict the current when a nanoparticle translocates through a single aperture. With their tiered three-dimensional structure, diatom shells can be used in to form nano-lipid bilayer membranes and can be used in ion-channel reconstitution experiments similar to synthetic nanoporous membranes.
To achieve a sensing diameter of 1-2 nanometers, the diatom shells were used as substrates to perform ion-channel reconstitution experiments. The immobilized diatom shell was functionalized using silane chemistry and lipid bilayer membranes were formed. Functionalization of the diatom shell surface improves bilayer formation probability from 1 out of 10 to 10 out of 10 as monitored by impedance spectroscopy. Self-insertion of outer membrane protein OmpF of E.Coli into the lipid membranes could be confirmed using single channel recordings, indicating that nano-BLMs had formed which allow for fully functional porin activity. The results indicate that biogenic silica nanoporous substrates can be simulated using a simplified two dimensional geometry to predict the current when a nanoparticle translocates through a single aperture. With their tiered three-dimensional structure, diatom shells can be used in to form nano-lipid bilayer membranes and can be used in ion-channel reconstitution experiments similar to synthetic nanoporous membranes.
ContributorsRamakrishnan, Shankar (Author) / Goryll, Michael (Thesis advisor) / Blain Christen, Jennifer (Committee member) / Dey, Sandwip (Committee member) / Thornton, Trevor (Committee member) / Arizona State University (Publisher)
Created2015
Description
Developing countries suffer from various health challenges due to inaccessible medical diagnostic laboratories and lack of resources to establish new laboratories. One way to address these issues is to develop diagnostic systems that are suitable for the low-resource setting. In addition to this, applications requiring rapid analyses further motivates the development of portable, easy-to-use, and accurate Point of Care (POC) diagnostics. Lateral Flow Immunoassays (LFIAs) are among the most successful POC tests as they satisfy most of the ASSURED criteria. However, factors like reagent stability, reaction rates limit the performance and robustness of LFIAs. The fluid flow rate in LFIA significantly affect the factors mentioned above, and hence, it is desirable to maintain an optimal fluid velocity in porous media.
The main objective of this study is to build a statistical model that enables us to determine the optimal design parameters and ambient conditions for achieving a desired fluid velocity in porous media. This study mainly focuses on the effects of relative humidity and temperature on evaporation in porous media and the impact of geometry on fluid velocity in LFIAs. A set of finite element analyses were performed, and the obtained simulation results were then experimentally verified using Whatman filter paper with different geometry under varying ambient conditions. Design of experiments was conducted to estimate the significant factors affecting the fluid flow rate.
Literature suggests that liquid evaporation is one of the major factors that inhibit fluid penetration and capillary flow in lateral flow Immunoassays. The obtained results closely align with the existing literature and conclude that a desired fluid flow rate can be achieved by tuning the geometry of the porous media. The derived statistical model suggests that a dry and warm atmosphere is expected to inhibit the fluid flow rate the most and vice-versa.
The main objective of this study is to build a statistical model that enables us to determine the optimal design parameters and ambient conditions for achieving a desired fluid velocity in porous media. This study mainly focuses on the effects of relative humidity and temperature on evaporation in porous media and the impact of geometry on fluid velocity in LFIAs. A set of finite element analyses were performed, and the obtained simulation results were then experimentally verified using Whatman filter paper with different geometry under varying ambient conditions. Design of experiments was conducted to estimate the significant factors affecting the fluid flow rate.
Literature suggests that liquid evaporation is one of the major factors that inhibit fluid penetration and capillary flow in lateral flow Immunoassays. The obtained results closely align with the existing literature and conclude that a desired fluid flow rate can be achieved by tuning the geometry of the porous media. The derived statistical model suggests that a dry and warm atmosphere is expected to inhibit the fluid flow rate the most and vice-versa.
ContributorsThamatam, Nipun (Author) / Christen, Jennifer Blain (Thesis advisor) / Goryll, Michael (Committee member) / Thornton, Trevor (Committee member) / Arizona State University (Publisher)
Created2019