Matching Items (1,583)
Filtering by
- All Subjects: Computer Science
- All Subjects: Mechanical Engineering
- Status: Published
Description
Many real-world planning problems can be modeled as Markov Decision Processes (MDPs) which provide a framework for handling uncertainty in outcomes of action executions. A solution to such a planning problem is a policy that handles possible contingencies that could arise during execution. MDP solvers typically construct policies for a problem instance without re-using information from previously solved instances. Research in generalized planning has demonstrated the utility of constructing algorithm-like plans that reuse such information. However, using such techniques in an MDP setting has not been adequately explored.
This thesis presents a novel approach for learning generalized partial policies that can be used to solve problems with different object names and/or object quantities using very few example policies for learning. This approach uses abstraction for state representation, which allows the identification of patterns in solutions such as loops that are agnostic to problem-specific properties. This thesis also presents some theoretical results related to the uniqueness and succinctness of the policies computed using such a representation. The presented algorithm can be used as fast, yet greedy and incomplete method for policy computation while falling back to a complete policy search algorithm when needed. Extensive empirical evaluation on discrete MDP benchmarks shows that this approach generalizes effectively and is often able to solve problems much faster than existing state-of-art discrete MDP solvers. Finally, the practical applicability of this approach is demonstrated by incorporating it in an anytime stochastic task and motion planning framework to successfully construct free-standing tower structures using Keva planks.
This thesis presents a novel approach for learning generalized partial policies that can be used to solve problems with different object names and/or object quantities using very few example policies for learning. This approach uses abstraction for state representation, which allows the identification of patterns in solutions such as loops that are agnostic to problem-specific properties. This thesis also presents some theoretical results related to the uniqueness and succinctness of the policies computed using such a representation. The presented algorithm can be used as fast, yet greedy and incomplete method for policy computation while falling back to a complete policy search algorithm when needed. Extensive empirical evaluation on discrete MDP benchmarks shows that this approach generalizes effectively and is often able to solve problems much faster than existing state-of-art discrete MDP solvers. Finally, the practical applicability of this approach is demonstrated by incorporating it in an anytime stochastic task and motion planning framework to successfully construct free-standing tower structures using Keva planks.
ContributorsKala Vasudevan, Deepak (Author) / Srivastava, Siddharth (Thesis advisor) / Zhang, Yu (Committee member) / Yang, Yezhou (Committee member) / Arizona State University (Publisher)
Created2020
Description
Next-generation sequencing is a powerful tool for detecting genetic variation. How-ever, it is also error-prone, with error rates that are much larger than mutation rates.
This can make mutation detection difficult; and while increasing sequencing depth
can often help, sequence-specific errors and other non-random biases cannot be de-
tected by increased depth. The problem of accurate genotyping is exacerbated when
there is not a reference genome or other auxiliary information available.
I explore several methods for sensitively detecting mutations in non-model or-
ganisms using an example Eucalyptus melliodora individual. I use the structure of
the tree to find bounds on its somatic mutation rate and evaluate several algorithms
for variant calling. I find that conventional methods are suitable if the genome of a
close relative can be adapted to the study organism. However, with structured data,
a likelihood framework that is aware of this structure is more accurate. I use the
techniques developed here to evaluate a reference-free variant calling algorithm.
I also use this data to evaluate a k-mer based base quality score recalibrator
(KBBQ), a tool I developed to recalibrate base quality scores attached to sequencing
data. Base quality scores can help detect errors in sequencing reads, but are often
inaccurate. The most popular method for correcting this issue requires a known
set of variant sites, which is unavailable in most cases. I simulate data and show
that errors in this set of variant sites can cause calibration errors. I then show that
KBBQ accurately recalibrates base quality scores while requiring no reference or other
information and performs as well as other methods.
Finally, I use the Eucalyptus data to investigate the impact of quality score calibra-
tion on the quality of output variant calls and show that improved base quality score
calibration increases the sensitivity and reduces the false positive rate of a variant
calling algorithm.
This can make mutation detection difficult; and while increasing sequencing depth
can often help, sequence-specific errors and other non-random biases cannot be de-
tected by increased depth. The problem of accurate genotyping is exacerbated when
there is not a reference genome or other auxiliary information available.
I explore several methods for sensitively detecting mutations in non-model or-
ganisms using an example Eucalyptus melliodora individual. I use the structure of
the tree to find bounds on its somatic mutation rate and evaluate several algorithms
for variant calling. I find that conventional methods are suitable if the genome of a
close relative can be adapted to the study organism. However, with structured data,
a likelihood framework that is aware of this structure is more accurate. I use the
techniques developed here to evaluate a reference-free variant calling algorithm.
I also use this data to evaluate a k-mer based base quality score recalibrator
(KBBQ), a tool I developed to recalibrate base quality scores attached to sequencing
data. Base quality scores can help detect errors in sequencing reads, but are often
inaccurate. The most popular method for correcting this issue requires a known
set of variant sites, which is unavailable in most cases. I simulate data and show
that errors in this set of variant sites can cause calibration errors. I then show that
KBBQ accurately recalibrates base quality scores while requiring no reference or other
information and performs as well as other methods.
Finally, I use the Eucalyptus data to investigate the impact of quality score calibra-
tion on the quality of output variant calls and show that improved base quality score
calibration increases the sensitivity and reduces the false positive rate of a variant
calling algorithm.
ContributorsOrr, Adam James (Author) / Cartwright, Reed (Thesis advisor) / Wilson, Melissa (Committee member) / Kusumi, Kenro (Committee member) / Taylor, Jesse (Committee member) / Pfeifer, Susanne (Committee member) / Arizona State University (Publisher)
Created2020
Description
Access to real-time situational information including the relative position and motion of surrounding objects is critical for safe and independent travel. Object or obstacle (OO) detection at a distance is primarily a task of the visual system due to the high resolution information the eyes are able to receive from afar. As a sensory organ in particular, the eyes have an unparalleled ability to adjust to varying degrees of light, color, and distance. Therefore, in the case of a non-visual traveler, someone who is blind or low vision, access to visual information is unattainable if it is positioned beyond the reach of the preferred mobility device or outside the path of travel. Although, the area of assistive technology in terms of electronic travel aids (ETA’s) has received considerable attention over the last two decades; surprisingly, the field has seen little work in the area focused on augmenting rather than replacing current non-visual travel techniques, methods, and tools. Consequently, this work describes the design of an intuitive tactile language and series of wearable tactile interfaces (the Haptic Chair, HaptWrap, and HapBack) to deliver real-time spatiotemporal data. The overall intuitiveness of the haptic mappings conveyed through the tactile interfaces are evaluated using a combination of absolute identification accuracy of a series of patterns and subjective feedback through post-experiment surveys. Two types of spatiotemporal representations are considered: static patterns representing object location at a single time instance, and dynamic patterns, added in the HaptWrap, which represent object movement over a time interval. Results support the viability of multi-dimensional haptics applied to the body to yield an intuitive understanding of dynamic interactions occurring around the navigator during travel. Lastly, it is important to point out that the guiding principle of this work centered on providing the navigator with spatial knowledge otherwise unattainable through current mobility techniques, methods, and tools, thus, providing the \emph{navigator} with the information necessary to make informed navigation decisions independently, at a distance.
ContributorsDuarte, Bryan Joiner (Author) / McDaniel, Troy (Thesis advisor) / Davulcu, Hasan (Committee member) / Li, Baoxin (Committee member) / Venkateswara, Hemanth (Committee member) / Arizona State University (Publisher)
Created2020
Description
Ultra High Performance (UHP) cementitious binders are a class of cement-based materials with high strength and ductility, designed for use in precast bridge connections, bridge superstructures, high load-bearing structural members like columns, and in structural repair and strengthening. This dissertation aims to elucidate the chemo-mechanical relationships in complex UHP binders to facilitate better microstructure-based design of these materials and develop machine learning (ML) models to predict their scale-relevant properties from microstructural information.To establish the connection between micromechanical properties and constitutive materials, nanoindentation and scanning electron microscopy experiments are performed on several cementitious pastes. Following Bayesian statistical clustering, mixed reaction products with scattered nanomechanical properties are observed, attributable to the low degree of reaction of the constituent particles, enhanced particle packing, and very low water-to-binder ratio of UHP binders. Relating the phase chemistry to the micromechanical properties, the chemical intensity ratios of Ca/Si and Al/Si are found to be important parameters influencing the incorporation of Al into the C-S-H gel.
ML algorithms for classification of cementitious phases are found to require only the intensities of Ca, Si, and Al as inputs to generate accurate predictions for more homogeneous cement pastes. When applied to more complex UHP systems, the overlapping chemical intensities in the three dominant phases – Ultra High Stiffness (UHS), unreacted cementitious replacements, and clinker – led to ML models misidentifying these three phases. Similarly, a reduced amount of data available on the hard and stiff UHS phases prevents accurate ML regression predictions of the microstructural phase stiffness using only chemical information. The use of generic virtual two-phase microstructures coupled with finite element analysis is also adopted to train MLs to predict composite mechanical properties. This approach applied to three different representations of composite materials produces accurate predictions, thus providing an avenue for image-based microstructural characterization of multi-phase composites such UHP binders. This thesis provides insights into the microstructure of the complex, heterogeneous UHP binders and the utilization of big-data methods such as ML to predict their properties. These results are expected to provide means for rational, first-principles design of UHP mixtures.
ML algorithms for classification of cementitious phases are found to require only the intensities of Ca, Si, and Al as inputs to generate accurate predictions for more homogeneous cement pastes. When applied to more complex UHP systems, the overlapping chemical intensities in the three dominant phases – Ultra High Stiffness (UHS), unreacted cementitious replacements, and clinker – led to ML models misidentifying these three phases. Similarly, a reduced amount of data available on the hard and stiff UHS phases prevents accurate ML regression predictions of the microstructural phase stiffness using only chemical information. The use of generic virtual two-phase microstructures coupled with finite element analysis is also adopted to train MLs to predict composite mechanical properties. This approach applied to three different representations of composite materials produces accurate predictions, thus providing an avenue for image-based microstructural characterization of multi-phase composites such UHP binders. This thesis provides insights into the microstructure of the complex, heterogeneous UHP binders and the utilization of big-data methods such as ML to predict their properties. These results are expected to provide means for rational, first-principles design of UHP mixtures.
ContributorsFord, Emily Lucile (Author) / Neithalath, Narayanan (Thesis advisor) / Rajan, Subramaniam D. (Committee member) / Mobasher, Barzin (Committee member) / Chawla, Nikhilesh (Committee member) / Hoover, Christian G. (Committee member) / Maneparambil, Kailas (Committee member) / Arizona State University (Publisher)
Created2020
Description
Water desalination has become one of the viable solutions to provide drinking water in regions with limited natural resources. This is particularly true in small communities in arid regions, which suffer from low rainfall, declining surface water and increasing salinity of groundwater. Yet, current desalination methods are difficult to be implemented in these areas due to their centralized large-scale design. In addition, these methods require intensive maintenance, and sometimes do not operate in high salinity feedwater. Membrane distillation (MD) is one technology that can potentially overcome these challenges and has received increasing attention in the last 15 years. The driving force of MD is the difference in vapor pressure across a microporous hydrophobic membrane. Compared to conventional membrane-based technologies, MD can treat high concentration feedwater, does not need intensive pretreatment, and has better fouling resistance. More importantly, MD operates at low feed temperatures and so it can utilize low–grade heat sources such as solar energy for its operation. While the integration of solar energy and MD was conventionally indirect (i.e. by having two separate systems: a solar collector and an MD module), recent efforts were focused on direct integration where the membrane itself is integrated within a solar collector aiming to have a more compact, standalone design suitable for small-scale applications. In this dissertation, a comprehensive review of these efforts is discussed in Chapter 2. Two novel direct solar-powered MD systems were proposed and investigated experimentally: firstly, a direct contact MD (DCMD) system was designed by placing capillary membranes within an evacuated tube solar collector (ETC) (Chapter 3), and secondly, a submerged vacuum MD (S-VMD) system that uses circulation and aeration as agitation techniques was investigated (Chapter 4). A maximum water production per absorbing area of 0.96 kg·m–2·h–1 and a thermal efficiency of 0.51 were achieved. A final study was conducted to investigate the effect of ultrasound in an S-VMD unit (Chapter 5), which significantly enhanced the permeate flux (up to 24%) and reduced the specific energy consumption (up to 14%). The results add substantially to the understanding of integrating ultrasound with different MD processes.
ContributorsBamasag, Ahmad (Author) / Phelan, Patrick E (Thesis advisor) / Shuaib, Abdelrahman (Committee member) / Wang, Liping (Committee member) / Bocanegra, Luis (Committee member) / Roedel, Ronald (Committee member) / Arizona State University (Publisher)
Created2020
Description
Over the past decade, machine learning research has made great strides and significant impact in several fields. Its success is greatly attributed to the development of effective machine learning algorithms like deep neural networks (a.k.a. deep learning), availability of large-scale databases and access to specialized hardware like Graphic Processing Units. When designing and training machine learning systems, researchers often assume access to large quantities of data that capture different possible variations. Variations in the data is needed to incorporate desired invariance and robustness properties in the machine learning system, especially in the case of deep learning algorithms. However, it is very difficult to gather such data in a real-world setting. For example, in certain medical/healthcare applications, it is very challenging to have access to data from all possible scenarios or with the necessary amount of variations as required to train the system. Additionally, the over-parameterized and unconstrained nature of deep neural networks can cause them to be poorly trained and in many cases over-confident which, in turn, can hamper their reliability and generalizability. This dissertation is a compendium of my research efforts to address the above challenges. I propose building invariant feature representations by wedding concepts from topological data analysis and Riemannian geometry, that automatically incorporate the desired invariance properties for different computer vision applications. I discuss how deep learning can be used to address some of the common challenges faced when working with topological data analysis methods. I describe alternative learning strategies based on unsupervised learning and transfer learning to address issues like dataset shifts and limited training data. Finally, I discuss my preliminary work on applying simple orthogonal constraints on deep learning feature representations to help develop more reliable and better calibrated models.
ContributorsSom, Anirudh (Author) / Turaga, Pavan (Thesis advisor) / Krishnamurthi, Narayanan (Committee member) / Spanias, Andreas (Committee member) / Li, Baoxin (Committee member) / Arizona State University (Publisher)
Created2020
Description
Deformable heat exchangers could provide a multitude of previously untapped advantages ranging from adaptable performance via macroscale, dynamic shape change (akin to dilation/constriction seen in blood vessels) to enhanced heat transfer at thermal interfaces through microscale, surface deformations. So far, making deformable, ‘soft heat exchangers’ (SHXs) has been limited by the low thermal conductivity of materials with suitable mechanical properties. The recent introduction of liquid-metal embedded elastomers by Bartlett et al1 has addressed this need. Specifically, by remaining soft and stretchable despite the addition of filler, these thermally conductive composites provide an ideal material for the new class of “soft thermal systems”, which is introduced in this work. Understanding such thermal systems will be a key element in enabling technology that require high levels of stretchability, such as thermoregulatory garments, soft electronics, wearable electronics, and high-powered robotics. Shape change inherent to SHX operation has the potential to violate many conventional assumptions used in HX design and thus requires the development of new theoretical approaches to predict performance. To create a basis for understanding these devices, this work highlights two sequential studies. First, the effects of transitioning to a surface deformable, SHX under steady state static conditions in the setting of a liquid cooling device for thermoregulation, electronics and robotics applications was explored. In this study, a thermomechanical model was built and validated to predict the thermal performance and a system wide analysis to optimize such devices was carried out. Second, from a more fundamental perspective, the effects of SHXs undergoing transient shape deformation during operation was explored. A phase shift phenomenon in cooling performance dependent on stretch rate, stretch extent and thermal diffusivity was discovered and explained. With the use of a time scale analysis, the extent of quasi-static assumption viability in modeling such systems was quantified and multiple shape modulation regime limits were defined. Finally, nuance considerations and future work of using liquid metal-silicone composites in SHXs were discussed.
ContributorsKotagama, Praveen (Author) / Rykaczewski, Konrad (Thesis advisor) / Wang, Robert (Committee member) / Phelan, Patrick (Committee member) / Herrmann, Marcus (Committee member) / Green, Matthew (Committee member) / Arizona State University (Publisher)
Created2020
Description
Video analysis and understanding have obtained more and more attention in recent years. The research community also has devoted considerable effort and made progress in many related visual tasks, like video action/event recognition, thumbnail frame or video index retrieval, and zero-shot learning. The way to find good representative features of videos is an important objective for these visual tasks.
Thanks to the success of deep neural networks in recent vision tasks, it is natural to take the deep learning methods into consideration for better extraction of a global representation of the images and videos. In general, Convolutional Neural Network (CNN) is utilized for obtaining the spatial information, and Recurrent Neural Network (RNN) is leveraged for capturing the temporal information.
This dissertation provides a perspective of the challenging problems in different kinds of videos which may require different solutions. Therefore, several novel deep learning-based approaches of obtaining representative features are outlined for different visual tasks like zero-shot learning, video retrieval, and video event recognition in this dissertation. To better understand and obtained the video spatial and temporal information, Convolutional Neural Network and Recurrent Neural Network are jointly utilized in most approaches. And different experiments are conducted to present the importance and effectiveness of good representative features for obtaining a better knowledge of video clips in the computer vision field. This dissertation also concludes a discussion with possible future works of obtaining better representative features of more challenging video clips.
Thanks to the success of deep neural networks in recent vision tasks, it is natural to take the deep learning methods into consideration for better extraction of a global representation of the images and videos. In general, Convolutional Neural Network (CNN) is utilized for obtaining the spatial information, and Recurrent Neural Network (RNN) is leveraged for capturing the temporal information.
This dissertation provides a perspective of the challenging problems in different kinds of videos which may require different solutions. Therefore, several novel deep learning-based approaches of obtaining representative features are outlined for different visual tasks like zero-shot learning, video retrieval, and video event recognition in this dissertation. To better understand and obtained the video spatial and temporal information, Convolutional Neural Network and Recurrent Neural Network are jointly utilized in most approaches. And different experiments are conducted to present the importance and effectiveness of good representative features for obtaining a better knowledge of video clips in the computer vision field. This dissertation also concludes a discussion with possible future works of obtaining better representative features of more challenging video clips.
ContributorsLi, Yikang (Author) / Li, Baoxin BL (Thesis advisor) / Karam, Lina LK (Committee member) / LiKamWa, Robert RL (Committee member) / Yang, Yezhou YY (Committee member) / Arizona State University (Publisher)
Created2020
Description
Most planning agents assume complete knowledge of the domain, which may not be the case in scenarios where certain domain knowledge is missing. This problem could be due to design flaws or arise from domain ramifications or qualifications. In such cases, planning algorithms could produce highly undesirable behaviors. Planning with incomplete domain knowledge is more challenging than partial observability in the sense that the planning agent is unaware of the existence of such knowledge, in contrast to it being just unobservable or partially observable. That is the difference between known unknowns and unknown unknowns.
In this thesis, I introduce and formulate this as the problem of Domain Concretization, which is inverse to domain abstraction studied extensively before. Furthermore, I present a solution that starts from the incomplete domain model provided to the agent by the designer and uses teacher traces from human users to determine the candidate model set under a minimalistic model assumption. A robust plan is then generated for the maximum probability of success under the set of candidate models. In addition to a standard search formulation in the model-space, I propose a sample-based search method and also an online version of it to improve search time. The solution presented has been evaluated on various International Planning Competition domains where incompleteness was introduced by deleting certain predicates from the complete domain model. The solution is also tested in a robot simulation domain to illustrate its effectiveness in handling incomplete domain knowledge. The results show that the plan generated by the algorithm increases the plan success rate without impacting action cost too much.
In this thesis, I introduce and formulate this as the problem of Domain Concretization, which is inverse to domain abstraction studied extensively before. Furthermore, I present a solution that starts from the incomplete domain model provided to the agent by the designer and uses teacher traces from human users to determine the candidate model set under a minimalistic model assumption. A robust plan is then generated for the maximum probability of success under the set of candidate models. In addition to a standard search formulation in the model-space, I propose a sample-based search method and also an online version of it to improve search time. The solution presented has been evaluated on various International Planning Competition domains where incompleteness was introduced by deleting certain predicates from the complete domain model. The solution is also tested in a robot simulation domain to illustrate its effectiveness in handling incomplete domain knowledge. The results show that the plan generated by the algorithm increases the plan success rate without impacting action cost too much.
ContributorsSharma, Akshay (Author) / Zhang, Yu (Thesis advisor) / Fainekos, Georgios (Committee member) / Srivastava, Siddharth (Committee member) / Arizona State University (Publisher)
Created2020
Description
Gas Dynamic Virtual Nozzles (GDVN) produce microscopic flow-focused liquid jets and are widely used for sample delivery in serial femtosecond crystallography (SFX) and time-resolved solution scattering. Recently, 2-photon polymerization (2PP) made it possible to produce 3D-printed GDVNs with submicron printing resolution. Comparing with hand- fabricated nozzles, reproducibility, and less developing effort, and similarity of the performance of different 3D printed nozzles are among the advantages of using 3D printing techniques to develop GDVN’s. Submicron printing resolution also makes it possible to easily improve GDVN performance by optimizing the design of nozzles. In this study, 3D printed nozzles were developed to achieve low liquid and gas flow rates and high liquid jet velocities. A double-pulsed nanosecond laser imaging system was used to perform Particle Tracking Velocimetry (PTV) in order to determine jet velocities and assess jet stability/reproducibility. The testing results of pure water jets focused with He sheath gas showed that some designs can easily achieve stable liquid jets with velocities of more than 80 m/s, with pure water flowing at 3 microliters/min, and helium sheath gas flowing at less than 5 mg/min respectively. A numerical simulation pipeline was also used to characterize the performance of different 3D printed GDVNs. The results highlight the potential of making reproducible GDVNs with minimum fabrication effort, that can meet the requirements of present and future SFX and time-resolved solution scattering research.
ContributorsNazari, Reza (Author) / Adrian, Ronald (Thesis advisor) / Kirian, Richard (Thesis advisor) / Herrmann, Marcus (Committee member) / Phelan, Patrick (Committee member) / Weierstall, Uwe (Committee member) / Arizona State University (Publisher)
Created2020