Computer Sciences Commons^™

A Study Of Machine Learning And Deep Learning Models For Solving Medical Imaging Problems, Fadi G. Farhat

Theses

Application of machine learning and deep learning methods on medical imaging aims to create systems that can help in the diagnosis of disease and the automation of analyzing medical images in order to facilitate treatment planning. Deep learning methods do well in image recognition, but medical images present unique challenges. The lack of large amounts of data, the image size, and the high class-imbalance in most datasets, makes training a machine learning model to recognize a particular pattern that is typically present only in case images a formidable task.

Experiments are conducted to classify breast cancer images as healthy or …

Deep Morphological Neural Networks, Yucong Shen

Theses

Mathematical morphology is a theory and technique applied to collect features like geometric and topological structures in digital images. Determining suitable morphological operations and structuring elements for a give purpose is a cumbersome and time-consuming task. In this paper, morphological neural networks are proposed to address this problem. Serving as a non-linear feature extracting layers in deep learning frameworks, the efficiency of the proposed morphological layer is confirmed analytically and empirically. With a known target, a single-filter morphological layer learns the structuring element correctly, and an adaptive layer can automatically select appropriate morphological operations. For high level applications, the proposed …

A Comparative Study Of Russian Trolls Using Several Machine Learning Models On Twitter Data, Kannan Neten Dharan Kannan Neten Dharan

Theses

Ever since Russian trolls have been brought into light, their interference in the 2016 US Presidential elections has been monitored and studied thoroughly. These Russian trolls have fake accounts registered on several major social media sites to influence public opinions. Our work involves trying to discover patterns in these tweets and classifying them by using different machine learning approaches such as Support Vector Machines, Word2vec and neural network models, and then creating a benchmark to compare all the different models. Two machine learning models are developed for this purpose. The first one is used to classify any given specific tweet …

Storage Systems For Mobile-Cloud Applications, Nafize R. Paiker

Dissertations

Mobile devices have become the major computing platform in todays world. However, some apps on mobile devices still suffer from insufficient computing and energy resources. A key solution is to offload resource-demanding computing tasks from mobile devices to the cloud. This leads to a scenario where computing tasks in the same application run concurrently on both the mobile device and the cloud.

This dissertation aims to ensure that the tasks in a mobile app that employs offloading can access and share files concurrently on the mobile and the cloud in a manner that is efficient, consistent, and transparent to locations. …

Deep Learning Methods For Mining Genomic Sequence Patterns, Xin Gao

Dissertations

Nowadays, with the growing availability of large-scale genomic datasets and advanced computational techniques, more and more data-driven computational methods have been developed to analyze genomic data and help to solve incompletely understood biological problems. Among them, deep learning methods, have been proposed to automatically learn and recognize the functional activity of DNA sequences from genomics data. Techniques for efficient mining genomic sequence pattern will help to improve our understanding of gene regulation, and thus accelerate our progress toward using personal genomes in medicine.

This dissertation focuses on the development of deep learning methods for mining genomic sequences. First, we compare …

Computational Intelligence In Steganography: Adaptive Image Watermarking, Xin Zhong

Dissertations

Digital image watermarking, as an extension of traditional steganography, refers to the process of hiding certain messages into cover images. The transport image, called marked-image or stego-image, conveys the hidden messages while appears visibly similar to the cover-image. Therefore, image watermarking enables various applications such as copyright protection and covert communication. In a watermarking scheme, fidelity, capacity and robustness are considered as crucial factors, where fidelity measures the similarity between the cover- and marked-images, capacity measures the maximum amount of watermark that can be embedded, and robustness concerns the watermark extraction under attacks on the marked-image. Watermarking techniques are often …

Polya Db3: A Database Cataloging Polyadenation Sites(Pas) Across Different Species And Their Conservation, Ram Mohan Nambiar

Theses

Polyadenation is an important process occurring in the messenger RNA that involves cleavage of 3 end nascent mRNAs and addition of poly(A) tails. For this thesis,I present PolyA DB3 ,a database cataloging cleavage and polyadenylation sites (PASs) in several genomes specifically for human,mouse,rat and chicken. This database is based on deep sequencing data. PASs are mapped by the 3’ region extraction and deep sequencing (3’READS) method, ensuring unequivocal PAS identification. Large volume of data based on diverse biological samples is used to increase PAS coverage and provide PAS usage information. Strand-specific RNA-seq data were used to extend annotated 3’ ends …

Improving K-Nn Search And Subspace Clustering Based On Local Intrinsic Dimensionality, Arwa M. Wali

Dissertations

In several novel applications such as multimedia and recommender systems, data is often represented as object feature vectors in high-dimensional spaces. The high-dimensional data is always a challenge for state-of-the-art algorithms, because of the so-called "curse of dimensionality". As the dimensionality increases, the discriminative ability of similarity measures diminishes to the point where many data analysis algorithms, such as similarity search and clustering, that depend on them lose their effectiveness. One way to handle this challenge is by selecting the most important features, which is essential for providing compact object representations as well as improving the overall search and clustering …

Applications Of Big Knowledge Summarization, Ling Zheng

Dissertations

Advanced technologies have resulted in the generation of large amounts of data ("Big Data"). The Big Knowledge derived from Big Data could be beyond humans' ability of comprehension, which will limit the effective and innovative use of Big Knowledge repository. Biomedical ontologies, which play important roles in biomedical information systems, constitute one kind of Big Knowledge repository. Biomedical ontologies typically consist of domain knowledge assertions expressed by the semantic connections between tens of thousands of concepts. Without some high-level visual representation of Big Knowledge in biomedical ontologies, humans cannot grasp the "big picture" of those ontologies. Such Big Knowledge orientation …

Novel Image Descriptors And Learning Methods For Image Classification Applications, Ajit Puthenputhussery

Dissertations

Image classification is an active and rapidly expanding research area in computer vision and machine learning due to its broad applications. With the advent of big data, the need for robust image descriptors and learning methods to process a large number of images for different kinds of visual applications has greatly increased. Towards that end, this dissertation focuses on exploring new image descriptors and learning methods by incorporating important visual aspects and enhancing the feature representation in the discriminative space for advancing image classification.

First, an innovative sparse representation model using the complete marginal Fisher analysis (CMFA-SR) framework is proposed …

Theoretical Studies Of Photoinduced Dynamics And Topological States In Materials With Strong Electron-Lattice Couplings, Linghua Zhu

Dissertations

First, we study the nonequilibrium dynamics of photoinduced phase transitions in charge ordered (CO) systems with a strong electron-lattice interaction and analyze the interplay between electrons, periodic lattice distortions, and a phonon thermal reservoir. Simulations based on a tight-binding Hamiltonian and Boltzmann equations reveal partially decoupled oscillations of the electronic order parameter and the periodic lattice distortion during CO melting, which becomes more energy efficient with lower photon energy. The cooling rate of the electron system correlates with the CO gap dynamics, responsible for an order of magnitude decrease of the cooling rate upon the gap reopening. The work also …

Virtual Smarts - Optimizing The Coalescing Of People For Collective Action Within Urban Communities, Stephen Thomas Ricken

Dissertations

Despite the importance of individuals coming together for social group-activities (e.g., pick-up volleyball), the process by which such groups coalesce is poorly understood, and as a consequence is poorly supported by technology. This is despite the emergence of Event-Based Social Network (EBSN) technologies that are specifically designed to assist group coalescing for social activities. Existing theories focus on group development in terms of norms and types, rather than the processes involved in initial group coalescence. This dissertation addresses this gap in the literature through four studies focusing on understanding the coalescing process for interest-based group activities within urban environments and …

High Performance Cloud Computing On Multicore Computers, Jianchen Shan

Dissertations

The cloud has become a major computing platform, with virtualization being a key to allow applications to run and share the resources in the cloud. A wide spectrum of applications need to process large amounts of data at high speeds in the cloud, e.g., analyzing customer data to find out purchase behavior, processing location data to determine geographical trends, or mining social media data to assess brand sentiment. To achieve high performance, these applications create and use multiple threads running on multicore processors. However, existing virtualization technology cannot support the efficient execution of such applications on virtual machines, making them …

Detecting And Characterizing Self Hiding Behavior In Android Applications, Raina Samuel

Theses

Applications (apps) that conceal their activities are fundamentally deceptive; app marketplaces and end-users should treat such apps as suspicious. However, due to its nature and intent, activity concealing is not disclosed up-front, which puts users at risk. This study focuses on characterization and detection of such techniques, e.g., hiding the app or removing traces, known as 'self hiding' (SH) behavior. SH behavior has not been studied per se - rather it has been reported on only as a byproduct of malware investigations. This gap is addressed via a study and suite of static analyses targeted at SH in Android apps. …

Hypoxic And Viral Contributions To The Etiopathogenesis Of Schizophrenia: A Whole Transcriptome Analysis, Kathryn A. Gorski

Theses

Schizophrenia is a mental illness with a complex and as of yet unclear etiology. It is highly heritable and has a strong polygenic character, however, studies examining the genetics of schizophrenia have not sufficiently explained all variability in its prevalence. Environmental causes are theorized to have a non trivial contribution to the pathoetiology of schizophrenia, including interactions with genetic components, but these mechanisms remain unclear. Analyzing schizophrenia dysfunction using transcriptomic approaches is a paradigm still in its infancy, and fewer studies still have examined non neurological contributions to schizophrenia pathology with next generation sequencing technologies. This pilot study uses several …

Supporting User Evaluation Of Messaging Interactions With Potential Romantic Partners Discovered Online, Douglas Zytko

Dissertations

Online dating systems have transformed the way people pursue romance. To arrive at a decision to meet for a face-to-face date, users gather information about each other online pertinent to romantic attraction. Yet sometimes they discover on the date that they made the wrong choice. One aspect of online dating system-use that may be a contributing factor, but is largely overlooked in the literature, is interaction through text-based messaging interfaces. This dissertation explores how messaging interactions inform face-to-face meeting decisions through two qualitative studies, and explores through a mixed methods field study how innovative messaging interfaces that embody theory from …

Characteristics Of Different Deep Neural Networks And Application Of Pre-Trained Model Without Transfer Learning, Zhiqi Peng

Theses

Deep neural networks have been successful in many areas, some of them even surpass human performances. The goal of this thesis is using data simulations to present different characteristics of three deep neural networks: fully connected deep neural network, convolutional neural network, recurrent neural network, which will perform best when dealing with different feature patterns. By using these characteristics to design a deep neural network on top of an adopted pre-trained model with untrainable layers, achieved an averagely 11.1% improvement than a model with transfer learning method.

Looping Predictive Method To Improve Accuracy Of A Machine Learning Model, Subramanyam Reddy Pogili

Theses

The topic of this project is an analysis of drug-related tweets. The goal is to build a Machine Learning Model that can distinguish between tweets that indicate drug abuse and other tweets that also contain the name of a drug but do not describe abuse. Drugs can be illegal, such as heroin, or legal drugs with a potential of abuse, such as painkillers. However, building a good Machine Learning Model requires a large amount of training data. For each training tweet, a human expert has determined whether it indicates drug abuse or not. This is difficult work for humans. …

A Conversation Centric Approach To Understanding And Supporting The Coordination Of Social Group-Activities, Richard P. Schuler

Dissertations

Despite the widespread and large variety of communication tools available to us such as, text messaging, Skype, email, twitter, Facebook, instant messaging, GroupMe, WhatsApp, Snapchat, etc., many people still routinely find coordinating activities with our friends to be a very frustrating experience. Everyone, has at least once, encountered the difficulties involved with deciding what to do as a group. Some friends may be busy, others may have already seen the movie that the others want to see, and some do not like Mexican food. It is a challenge everyone has faced and continue to face. This is a result of …

Annotation Of Multimedia Learning Materials For Semantic Search, Sheetal Rajgure

Dissertations

Multimedia is the main source for online learning materials, such as videos, slides and textbooks, and its size is growing with the popularity of online programs offered by Universities and Massive Open Online Courses (MOOCs). The increasing amount of multimedia learning resources available online makes it very challenging to browse through the materials or find where a specific concept of interest is covered. To enable semantic search on the lecture materials, their content must be annotated and indexed. Manual annotation of learning materials such as videos is tedious and cannot be envisioned for the growing quantity of online materials. One …

Programming Frameworks For Mobile Sensing, Hillol Debnath

Dissertations

The proliferation of smart mobile devices in people’s daily lives is making context-aware computing a reality. A plethora of sensors available in these devices can be utilized to understand users’ context better. Apps can provide more relevant data or services to the user based on improved understanding of user’s context. With the advent of cloud-assisted mobile platforms, apps can also perform collaborative computation over the sensing data collected from a group of users. However, there are still two main issues: (1) A lack of simple and effective personal sensing frameworks: existing frameworks do not provide support for real-time fusing of …

Matrix Completion Algorithms With Applications In Biomedicine, E-Commerce And Social Science, Yiran Wang

Theses

This thesis investigates matrix completion algorithms with applications in biomedicine, e-commerce and social science. In general, matrix completion algorithms work well for low rank matrices. Such matrices find many applications in recommender systems and social network analysis. On the other hand, biological networks often yield high rank matrices. For example, the adjacency matrix representing interactions between transcription factors and target genes in the cell is a highly sparse matrix, in which most entries correspond to absent interactions and only a few entries correspond to present interactions. This sparse matrix is a high rank or even full rank matrix. Matrix completion …

Detecting User Demographics In Twitter To Inform Health Trends In Social Media, Christopher R. Markson

Dissertations

The widespread and popular use of social media and social networking applications offer a promising opportunity for gaining knowledge and insights regarding population health conditions thanks to the diversity and abundance of online user-generated information (UGHI) relating to healthcare and well-being. However, users on social media and social networking sites often do not supply their complete demographic information, which greatly undermines the value of the aforementioned information for health 2.0 research, e.g., for discerning disparities across population groups in certain health conditions. To recover the missing user demographic information, existing methods observe a limited scope of user behaviors, such as …

Performance Optimization And Energy Efficiency Of Big-Data Computing Workflows, Tong Shu

Dissertations

Next-generation e-science is producing colossal amounts of data, now frequently termed as Big Data, on the order of terabyte at present and petabyte or even exabyte in the predictable future. These scientific applications typically feature data-intensive workflows comprised of moldable parallel computing jobs, such as MapReduce, with intricate inter-job dependencies. The granularity of task partitioning in each moldable job of such big data workflows has a significant impact on workflow completion time, energy consumption, and financial cost if executed in clouds, which remains largely unexplored. This dissertation conducts an in-depth investigation into the properties of moldable jobs and provides an …

Development And Evaluation Of Machine Learning Algorithms For Biomedical Applications, Turki Talal Turki

Dissertations

Gene network inference and drug response prediction are two important problems in computational biomedicine. The former helps scientists better understand the functional elements and regulatory circuits of cells. The latter helps a physician gain full understanding of the effective treatment on patients. Both problems have been widely studied, though current solutions are far from perfect. More research is needed to improve the accuracy of existing approaches.

This dissertation develops machine learning and data mining algorithms, and applies these algorithms to solve the two important biomedical problems. Specifically, to tackle the gene network inference problem, the dissertation proposes (i) new techniques …

Big Data Analytics In Computational Biology And Bioinformatics, Kevin Byron

Dissertations

Big data analytics in computational biology and bioinformatics refers to an array of operations including biological pattern discovery, classification, prediction, inference, clustering as well as data mining in the cloud, among others. This dissertation addresses big data analytics by investigating two important operations, namely pattern discovery and network inference.

The dissertation starts by focusing on biological pattern discovery at a genomic scale. Research reveals that the secondary structure in non-coding RNA (ncRNA) is more conserved during evolution than its primary nucleotide sequence. Using a covariance model approach, the stems and loops of an ncRNA secondary structure are represented as a …

Viewability Prediction For Display Advertising, Chong Wang

Dissertations

As a massive industry, display advertising delivers advertisers’ marketing messages to attract customers through graphic banners on webpages. Display advertising is also the most essential revenue source of online publishers. Currently, advertisers are charged by user response or ad serving. However, recent studies show that users barely click or convert display ads. Moreover, about half of the ads are actually never seen by users. In this case, advertisers cannot enhance their brand awareness and increase return on investment. Publishers also lose much revenue. Therefore, the ad pricing standards are shifting to a new model: ad impressions are paid if they …

Investigation Of New Learning Methods For Visual Recognition, Qingfeng Liu

Dissertations

Visual recognition is one of the most difficult and prevailing problems in computer vision and pattern recognition due to the challenges in understanding the semantics and contents of digital images. Two major components of a visual recognition system are discriminatory feature representation and efficient and accurate pattern classification. This dissertation therefore focuses on developing new learning methods for visual recognition.

Based on the conventional sparse representation, which shows its robustness for visual recognition problems, a series of new methods is proposed. Specifically, first, a new locally linear K nearest neighbor method, or LLK method, is presented. The LLK method derives …

Statistical Learning Methods For Mining Marketing And Biological Data, Jie Zhang

Dissertations

Nowadays, the value of data has been broadly recognized and emphasized. More and more decisions are made based on data and analysis rather than solely on experience and intuition. With the fast development of networking, data storage, and data collection capacity, data have increased dramatically in industry, science and engineering domains, which brings both great opportunities and challenges. To take advantage of the data flood, new computational methods are in demand to process, analyze and understand these datasets.

This dissertation focuses on the development of statistical learning methods for online advertising and bioinformatics to model real world data with temporal …

Algorithms For Pre-Microrna Classification And A Gpu Program For Whole Genome Comparison, Ling Zhong

Dissertations

MicroRNAs (miRNAs) are non-coding RNAs with approximately 22 nucleotides that are derived from precursor molecules. These precursor molecules or pre-miRNAs often fold into stem-loop hairpin structures. However, a large number of sequences with pre-miRNA-like hairpin can be found in genomes. It is a challenge to distinguish the real pre-miRNAs from other hairpin sequences with similar stem-loops (referred to as pseudo pre-miRNAs). The first part of this dissertation presents a new method, called MirID, for identifying and classifying microRNA precursors. MirID is comprised of three steps. Initially, a combinatorial feature mining algorithm is developed to identify suitable feature sets. Then, the …