Matching Items (75)
Filtering by
- All Subjects: Mechanical Engineering
- All Subjects: Machine Learning
- Creators: Herrmann, Marcus
- Creators: Li, Baoxin
- Member of: ASU Electronic Theses and Dissertations
- Status: Published
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
In recent years, machine learning and data mining technologies have received growing attention in several areas such as recommendation systems, natural language processing, speech and handwriting recognition, image processing and biomedical domain. Many of these applications which deal with physiological and biomedical data require person specific or person adaptive systems. The greatest challenge in developing such systems is the subject-dependent data variations or subject-based variability in physiological and biomedical data, which leads to difference in data distributions making the task of modeling these data, using traditional machine learning algorithms, complex and challenging. As a result, despite the wide application of machine learning, efficient deployment of its principles to model real-world data is still a challenge. This dissertation addresses the problem of subject based variability in physiological and biomedical data and proposes person adaptive prediction models based on novel transfer and active learning algorithms, an emerging field in machine learning. One of the significant contributions of this dissertation is a person adaptive method, for early detection of muscle fatigue using Surface Electromyogram signals, based on a new multi-source transfer learning algorithm. This dissertation also proposes a subject-independent algorithm for grading the progression of muscle fatigue from 0 to 1 level in a test subject, during isometric or dynamic contractions, at real-time. Besides subject based variability, biomedical image data also varies due to variations in their imaging techniques, leading to distribution differences between the image databases. Hence a classifier learned on one database may perform poorly on the other database. Another significant contribution of this dissertation has been the design and development of an efficient biomedical image data annotation framework, based on a novel combination of transfer learning and a new batch-mode active learning method, capable of addressing the distribution differences across databases. The methodologies developed in this dissertation are relevant and applicable to a large set of computing problems where there is a high variation of data between subjects or sources, such as face detection, pose detection and speech recognition. From a broader perspective, these frameworks can be viewed as a first step towards design of automated adaptive systems for real world data.
ContributorsChattopadhyay, Rita (Author) / Panchanathan, Sethuraman (Thesis advisor) / Ye, Jieping (Thesis advisor) / Li, Baoxin (Committee member) / Santello, Marco (Committee member) / Arizona State University (Publisher)
Created2013
Description
The atomization of a liquid jet by a high speed cross-flowing gas has many applications such as gas turbines and augmentors. The mechanisms by which the liquid jet initially breaks up, however, are not well understood. Experimental studies suggest the dependence of spray properties on operating conditions and nozzle geom- etry. Detailed numerical simulations can offer better understanding of the underlying physical mechanisms that lead to the breakup of the injected liquid jet. In this work, detailed numerical simulation results of turbulent liquid jets injected into turbulent gaseous cross flows for different density ratios is presented. A finite volume, balanced force fractional step flow solver to solve the Navier-Stokes equations is employed and coupled to a Refined Level Set Grid method to follow the phase interface. To enable the simulation of atomization of high density ratio fluids, we ensure discrete consistency between the solution of the conservative momentum equation and the level set based continuity equation by employing the Consistent Rescaled Momentum Transport (CRMT) method. The impact of different inflow jet boundary conditions on different jet properties including jet penetration is analyzed and results are compared to those obtained experimentally by Brown & McDonell(2006). In addition, instability analysis is performed to find the most dominant insta- bility mechanism that causes the liquid jet to breakup. Linear instability analysis is achieved using linear theories for Rayleigh-Taylor and Kelvin- Helmholtz instabilities and non-linear analysis is performed using our flow solver with different inflow jet boundary conditions.
ContributorsGhods, Sina (Author) / Herrmann, Marcus (Thesis advisor) / Squires, Kyle (Committee member) / Chen, Kangping (Committee member) / Huang, Huei-Ping (Committee member) / Tang, Wenbo (Committee member) / Arizona State University (Publisher)
Created2013
Description
We solve the problem of activity verification in the context of sustainability. Activity verification is the process of proving the user assertions pertaining to a certain activity performed by the user. Our motivation lies in incentivizing the user for engaging in sustainable activities like taking public transport or recycling. Such incentivization schemes require the system to verify the claim made by the user. The system verifies these claims by analyzing the supporting evidence captured by the user while performing the activity. The proliferation of portable smart-phones in the past few years has provided us with a ubiquitous and relatively cheap platform, having multiple sensors like accelerometer, gyroscope, microphone etc. to capture this evidence data in-situ. In this research, we investigate the supervised and semi-supervised learning techniques for activity verification. Both these techniques make use the data set constructed using the evidence submitted by the user. Supervised learning makes use of annotated evidence data to build a function to predict the class labels of the unlabeled data points. The evidence data captured can be either unimodal or multimodal in nature. We use the accelerometer data as evidence for transportation mode verification and image data as evidence for recycling verification. After training the system, we achieve maximum accuracy of 94% when classifying the transport mode and 81% when detecting recycle activity. In the case of recycle verification, we could improve the classification accuracy by asking the user for more evidence. We present some techniques to ask the user for the next best piece of evidence that maximizes the probability of classification. Using these techniques for detecting recycle activity, the accuracy increases to 93%. The major disadvantage of using supervised models is that it requires extensive annotated training data, which expensive to collect. Due to the limited training data, we look at the graph based inductive semi-supervised learning methods to propagate the labels among the unlabeled samples. In the semi-supervised approach, we represent each instance in the data set as a node in the graph. Since it is a complete graph, edges interconnect these nodes, with each edge having some weight representing the similarity between the points. We propagate the labels in this graph, based on the proximity of the data points to the labeled nodes. We estimate the performance of these algorithms by measuring how close the probability distribution of the data after label propagation is to the probability distribution of the ground truth data. Since labeling has a cost associated with it, in this thesis we propose two algorithms that help us in selecting minimum number of labeled points to propagate the labels accurately. Our proposed algorithm achieves a maximum of 73% increase in performance when compared to the baseline algorithm.
ContributorsDesai, Vaishnav (Author) / Sundaram, Hari (Thesis advisor) / Li, Baoxin (Thesis advisor) / Turaga, Pavan (Committee member) / Arizona State University (Publisher)
Created2013
Description
The heat transfer enhancements available from expanding the cross-section of a boiling microchannel are explored analytically and experimentally. Evaluation of the literature on critical heat flux in flow boiling and associated pressure drop behavior is presented with predictive critical heat flux (CHF) and pressure drop correlations. An optimum channel configuration allowing maximum CHF while reducing pressure drop is sought. A perturbation of the channel diameter is employed to examine CHF and pressure drop relationships from the literature with the aim of identifying those adequately general and suitable for use in a scenario with an expanding channel. Several CHF criteria are identified which predict an optimizable channel expansion, though many do not. Pressure drop relationships admit improvement with expansion, and no optimum presents itself. The relevant physical phenomena surrounding flow boiling pressure drop are considered, and a balance of dimensionless numbers is presented that may be of qualitative use. The design, fabrication, inspection, and experimental evaluation of four copper microchannel arrays of different channel expansion rates with R-134a refrigerant is presented. Optimum rates of expansion which maximize the critical heat flux are considered at multiple flow rates, and experimental results are presented demonstrating optima. The effect of expansion on the boiling number is considered, and experiments demonstrate that expansion produces a notable increase in the boiling number in the region explored, though no optima are observed. Significant decrease in the pressure drop across the evaporator is observed with the expanding channels, and no optima appear. Discussion of the significance of this finding is presented, along with possible avenues for future work.
ContributorsMiner, Mark (Author) / Phelan, Patrick E (Thesis advisor) / Baer, Steven (Committee member) / Chamberlin, Ralph (Committee member) / Chen, Kangping (Committee member) / Herrmann, Marcus (Committee member) / Arizona State University (Publisher)
Created2013
Description
Increasing computational demands in data centers require facilities to operate at higher ambient temperatures and at higher power densities. Conventionally, data centers are cooled with electrically-driven vapor-compressor equipment. This paper proposes an alternative data center cooling architecture that is heat-driven. The source is heat produced by the computer equipment. This dissertation details experiments investigating the quantity and quality of heat that can be captured from a liquid-cooled microprocessor on a computer server blade from a data center. The experiments involve four liquid-cooling setups and associated heat-extraction, including a radical approach using mineral oil. The trials examine the feasibility of using the thermal energy from a CPU to drive a cooling process. Uniquely, the investigation establishes an interesting and useful relationship simultaneously among CPU temperatures, power, and utilization levels. In response to the system data, this project explores the heat, temperature and power effects of adding insulation, varying water flow, CPU loading, and varying the cold plate-to-CPU clamping pressure. The idea is to provide an optimal and steady range of temperatures necessary for a chiller to operate. Results indicate an increasing relationship among CPU temperature, power and utilization. Since the dissipated heat can be captured and removed from the system for reuse elsewhere, the need for electricity-consuming computer fans is eliminated. Thermocouple readings of CPU temperatures as high as 93°C and a calculated CPU thermal energy up to 67Wth show a sufficiently high temperature and thermal energy to serve as the input temperature and heat medium input to an absorption chiller. This dissertation performs a detailed analysis of the exergy of a processor and determines the maximum amount of energy utilizable for work. Exergy as a source of realizable work is separated into its two contributing constituents: thermal exergy and informational exergy. The informational exergy is that usable form of work contained within the most fundamental unit of information output by a switching device within a CPU. Exergetic thermal, informational and efficiency values are calculated and plotted for our particular CPU, showing how the datasheet standards compare with experimental values. The dissertation concludes with a discussion of the work's significance.
ContributorsHaywood, Anna (Author) / Phelan, Patrick E (Thesis advisor) / Herrmann, Marcus (Committee member) / Gupta, Sandeep (Committee member) / Trimble, Steve (Committee member) / Myhajlenko, Stefan (Committee member) / Arizona State University (Publisher)
Created2014
Description
Surgery as a profession requires significant training to improve both clinical decision making and psychomotor proficiency. In the medical knowledge domain, tools have been developed, validated, and accepted for evaluation of surgeons' competencies. However, assessment of the psychomotor skills still relies on the Halstedian model of apprenticeship, wherein surgeons are observed during residency for judgment of their skills. Although the value of this method of skills assessment cannot be ignored, novel methodologies of objective skills assessment need to be designed, developed, and evaluated that augment the traditional approach. Several sensor-based systems have been developed to measure a user's skill quantitatively, but use of sensors could interfere with skill execution and thus limit the potential for evaluating real-life surgery. However, having a method to judge skills automatically in real-life conditions should be the ultimate goal, since only with such features that a system would be widely adopted. This research proposes a novel video-based approach for observing surgeons' hand and surgical tool movements in minimally invasive surgical training exercises as well as during laparoscopic surgery. Because our system does not require surgeons to wear special sensors, it has the distinct advantage over alternatives of offering skills assessment in both learning and real-life environments. The system automatically detects major skill-measuring features from surgical task videos using a computing system composed of a series of computer vision algorithms and provides on-screen real-time performance feedback for more efficient skill learning. Finally, the machine-learning approach is used to develop an observer-independent composite scoring model through objective and quantitative measurement of surgical skills. To increase effectiveness and usability of the developed system, it is integrated with a cloud-based tool, which automatically assesses surgical videos upload to the cloud.
ContributorsIslam, Gazi (Author) / Li, Baoxin (Thesis advisor) / Liang, Jianming (Thesis advisor) / Dinu, Valentin (Committee member) / Greenes, Robert (Committee member) / Smith, Marshall (Committee member) / Kahol, Kanav (Committee member) / Patel, Vimla L. (Committee member) / Arizona State University (Publisher)
Created2013
Description
The objective of this research is to develop methods for generating the Tolerance-Map for a line-profile that is specified by a designer to control the geometric profile shape of a surface. After development, the aim is to find one that can be easily implemented in computer software using existing libraries. Two methods were explored: the parametric modeling method and the decomposed modeling method. The Tolerance-Map (T-Map) is a hypothetical point-space, each point of which represents one geometric variation of a feature in its tolerance-zone. T-Maps have been produced for most of the tolerance classes that are used by designers, but, prior to the work of this project, the method of construction required considerable intuitive input, rather than being based primarily on automated computer tools. Tolerances on line-profiles are used to control cross-sectional shapes of parts, such as every cross-section of a mildly twisted compressor blade. Such tolerances constrain geometric manufacturing variations within a specified two-dimensional tolerance-zone. A single profile tolerance may be used to control position, orientation, and form of the cross-section. Four independent variables capture all of the profile deviations: two independent translations in the plane of the profile, one rotation in that plane, and the size-increment necessary to identify one of the allowable parallel profiles. For the selected method of generation, the line profile is decomposed into three types of segments, a primitive T-Map is produced for each segment, and finally the T-Maps from all the segments are combined to obtain the T-Map for the given profile. The types of segments are the (straight) line-segment, circular arc-segment, and the freeform-curve segment. The primitive T-Maps are generated analytically, and, for freeform-curves, they are built approximately with the aid of the computer. A deformation matrix is used to transform the primitive T-Maps to a single coordinate system for the whole profile. The T-Map for the whole line profile is generated by the Boolean intersection of the primitive T-Maps for the individual profile segments. This computer-implemented method can generate T-Maps for open profiles, closed ones, and those containing concave shapes.
ContributorsHe, Yifei (Author) / Davidson, Joseph (Thesis advisor) / Shah, Jami (Committee member) / Herrmann, Marcus (Committee member) / Arizona State University (Publisher)
Created2013
Description
Multi-pulse particle tracking velocimetry (multi-pulse PTV) is a recently proposed flow measurement technique aiming to improve the performance of conventional PTV/ PIV. In this work, multi-pulse PTV is assessed based on PTV simulations in terms of spatial resolution, velocity measurement accuracy and the capability of acceleration measurement. The errors of locating particles, velocity measurement and acceleration measurement are analytically calculated and compared among quadruple-pulse, triple-pulse and dual-pulse PTV. The optimizations of triple-pulse and quadruple-pulse PTV are discussed, and criteria are developed to minimize the combined error in position, velocity and acceleration. Experimentally, the velocity and acceleration fields of a round impinging air jet are measured to test the triple-pulse technique. A high speed beam-splitting camera and a custom 8-pulsed laser system are utilized to achieve good timing flexibility and temporal resolution. A new method to correct the registration error between CCDs is also presented. Consequently, the velocity field shows good consistency between triple-pulse and dual-pulse measurements. The mean acceleration profile along the centerline of the jet is used as the ground truth for the verification of the triple-pulse PIV measurements of the acceleration fields. The instantaneous acceleration field of the jet is directly measured by triple-pulse PIV and presented. Accelerations up to 1,000 g's are measured in these experiments.
ContributorsDing, Liuyang (Author) / Adrian, Ronald J. (Thesis advisor) / Herrmann, Marcus (Committee member) / Huang, Huei-Ping (Committee member) / Arizona State University (Publisher)
Created2014
Description
In many fields one needs to build predictive models for a set of related machine learning tasks, such as information retrieval, computer vision and biomedical informatics. Traditionally these tasks are treated independently and the inference is done separately for each task, which ignores important connections among the tasks. Multi-task learning aims at simultaneously building models for all tasks in order to improve the generalization performance, leveraging inherent relatedness of these tasks. In this thesis, I firstly propose a clustered multi-task learning (CMTL) formulation, which simultaneously learns task models and performs task clustering. I provide theoretical analysis to establish the equivalence between the CMTL formulation and the alternating structure optimization, which learns a shared low-dimensional hypothesis space for different tasks. Then I present two real-world biomedical informatics applications which can benefit from multi-task learning. In the first application, I study the disease progression problem and present multi-task learning formulations for disease progression. In the formulations, the prediction at each point is a regression task and multiple tasks at different time points are learned simultaneously, leveraging the temporal smoothness among the tasks. The proposed formulations have been tested extensively on predicting the progression of the Alzheimer's disease, and experimental results demonstrate the effectiveness of the proposed models. In the second application, I present a novel data-driven framework for densifying the electronic medical records (EMR) to overcome the sparsity problem in predictive modeling using EMR. The densification of each patient is a learning task, and the proposed algorithm simultaneously densify all patients. As such, the densification of one patient leverages useful information from other patients.
ContributorsZhou, Jiayu (Author) / Ye, Jieping (Thesis advisor) / Mittelmann, Hans (Committee member) / Li, Baoxin (Committee member) / Wang, Yalin (Committee member) / Arizona State University (Publisher)
Created2014