91 results
Search Results
Now showing 1 - 10 of 91
Item Context Specific Module Mining from Multiple Co-Expression Graph(North Dakota State University, 2017) Hossain, Md ShakhawatGene co-expression networks can be used to associate genes of unknown function with biological processes or to find genes in a specific context, environment responsible for a disease. We provide an overview of methods and tools used to identify such recurrent patterns across multiple networks, can be used to discover biological modules in co-expression networks constructed from gene expression data and we explain how this can be used to identify genes with a regulatory role in disease. However, existing algorithms are very much costly in terms of time and space. As network size or number increases, mining such modules get much more complex. We have developed an efficient approach to mine such recurrent context specific modules from 35 gene networks. This computationally very difficult problem due to the exponential number of patterns was solved non-exponentially.Item Sentiment Analysis and Opinion Mining on Twitter with GMO Keyword(North Dakota State University, 2016) Li, HanzheTwitter are a new source of information for data mining techniques. Messages posted through Twitter provide a major information source to gauge public sentiment on topics ranging from politics to fashion trends. The purpose of this paper is to analyze the Twitter tweets to discern the opinions of users regarding Genetically Modified Organisms (GMOs). We examine the effectiveness of several classifiers, Multinomial Naïve Bayes, Bernoulli Naïve Bayes, Logistic Regression and Linear Support Vector Classifier (SVC) in identifying a positive, negative or neutral category on a tweet corpus. Additionally, we use three datasets in this experiment to examine which dataset has the best score. Comparing the classifiers, we discovered that GMO_NDSU has the highest score in each classifier of my experiment among three datasets, and Linear SVC had the highest consistent accuracy by using bigrams as feature extraction and Term Frequency, Chi Square as feature selection.Item Deep Learning Applied to Public Company Valuation for Value Investing(North Dakota State University, 2021) Haich, Abram PaulValue investing is an investing approach that seeks to discover and take advantage of price discrepancies between the market price and the actual value of a company (intrinsic value). The purpose of this work is to measure the intrinsic value of companies using an approach that has had success in the broad field of Artificial Intelligence, Deep Learning. Finding patterns in large amounts of data is what Deep Learning can be used for. Typically for value investing an investor will seek to find conservative estimates on the current value of a company by analyzing fundamental data. Our method attempts to perform these estimates in a data driven manor using Deep Learning to estimate the intrinsic value of a company with the overall goal of aiding the Investor in uncovering undervalued companies.Item Identifying & Analyzing Security Vulnerabilities While Integration of Data from Cloud to SQL DBMS An Industrial Case Study(North Dakota State University, 2020) Kaur, SimranjitData integration defines propulsion of data from numerous sources into a database system in a way that yields the information representing powerful analytics. Various industries are always working towards keeping their data confidential and in a manner that will let them mine it later & retrieve strong statistics out of it for procuring future profits. This research will discuss about using SaaS (Software as a service) and SQL (Structured Query Language) as a combined model for storing inspections data to achieve the above-mentioned goals that the companies are on the road to. Since SaaS and SQL comes into play, a major part of the research automatically pops up, i.e. security vulnerabilities. Hence, this study details about various security threats while using Cloud storage, SQL database and during the ETL (Extraction, Transformation & Loading) from former to latter, in addition to connecting these database systems to produce an overall secure system.Item Athletic Fundraising and University Development Offices: A Structurational Relationship(North Dakota State University, 2010) Dickhudt, Keith MichaelThis case study examines the working relationship between an athletic department and a central development office within a university. This study focuses primarily on the coordination of fundraising efforts between the two offices. A qualitative approach, using a structuration theoretical framework, presents the working relationship at Midwestern State University (MWSU) through in-depth interviews. Results suggest the two offices could improve the coordination of fundraising efforts. Recommendations, based on the results and theoretical framework, are given.Item Application of Memory-Based Collaborative Filtering to Predict Fantasy Points of NFL Quarterbacks(North Dakota State University, 2019) Paramarta, Dienul Haq AmbegSubjective expert projections have been traditionally used to predict points in fantasy football, while machine prediction applications are limited. Memory-based collaborative filtering has been widely used in recommender system domain to predict ratings and recommend items. In this study, user-based and item-based collaborative filtering were explored and implemented to predict the weekly statistics and fantasy points of NFL quarterbacks. The predictions from three seasons were compared against expert projections. On both weekly statistics and total fantasy points, the implementations could not make significantly better predictions than experts.However, the prediction from the implementation improved the accuracy of other regression models when used as additional feature.Item Smart Grid Optimization Using a Capacitated Transshipment Problem Solver(North Dakota State University, 2013) Lampl, DamianA network flow model known as the capacitated transshipment problem, or CTP, can represent key aspects of a smart grid test network with the goal of finding minimum cost electric power flows using multiple different cost performance metrics. A custom CTP Solver was developed and implemented as an ASP.NET web application in an effort to study these various minimum cost smart grid problems and provide their optimal solutions. The CTP Solver modifies traditional linear programming concepts by introducing object oriented software development practices, as well as an insightful innovation for handling bidirectional arcs, which effectively halves the required disk and memory allocation of fully bidirectional networks. As an initial step toward smart grid optimization problem solutions, the CTP Solver provides a glimpse of how self-healing and possibly other key components of smart grid architecture might be handled in the future.Item Efficient Publish/Subscribe System over Mobile Ad-Hoc Network(North Dakota State University, 2012) Liu, ChaoInformation dissemination is an important issue for mobile ad-hoc communities. This issue is very challenging due to the dynamic and fragile nature of the mobile ad-hoc networks, in which participants have limited computing resources and battery, intermittent network connections, and mobile tasks. To address the aforementioned issue, this thesis proposes an efficient semantics-based publish/subscribe strategy. In our proposed publish/subscribe system, distributed mobile participants are organized into clusters based on their location proximity. A compact semantics-based indexing scheme is provided to guide information flow. Intra- and inter- cluster routings are proposed to assist efficient propagation of event notifications. A comprehensive set of simulation experiments prove the effectiveness of the proposed scheme.Item A Restful Architecture for Multiuser Virtual Environments and Simulations(North Dakota State University, 2014) Kariluoma, MattiJavaMOO is an architecture for creating multiuser virtual environments using the MUD (Multi-User Dungeon) and MOO (MUD Object Oriented) design patterns (rooms and objects in the rooms, including \exit" objects that lead to other rooms). The MOO design pattern traces it roots back to the rst multiuser virtual environments in the early 1980s. The focus of this thesis is joining the MOO design pattern with a distributed architecture. A distributed architecture is pursued to reduce the per-server computational load, compared to a traditional single-server approach. The e ect of transferring computational load to the client software is also investigated, with particular attention to the case of a graphical client with rich 3D visualizations. This results in an architecture employing a RESTful (REpresentational State Transfer) common interface and non-authoritative state synchronization that supports the MOO design pattern, uses fewer server-side resources, and is deployable as a network of distributed servers.Item Mining Novel Knowledge from Biomedical Literature using Statistical Measures and Domain Knowledge(North Dakota State University, 2016) Jha, KishlayThe problem of inferring novel knowledge from implicit facts by logically connecting independent fragments of literature is known as Literature Based Discovery (LBD). In LBD, to discover hidden links, it is important to determine the relevancy between concepts using appropriate information measures. In this study, to discover interesting and inherent links latent in large corpora, nine distinct methods, comprising variants of statistical information measures and derived semantic knowledge from domain ontology, are designed and compared. A series of experiments are performed and analyzed for those proposed methods. Also, a new strategy of effective preprocessing is proposed, which is capable of removing terms that have meager chances of constituting a new discovery. Finally, an organized list of final concepts deemed worthy of scientific investigation are provided to the user. Overall, our research presents a comprehensive analysis and perspective of how different statistical information measures and semantic knowledge affect the knowledge discovery procedure.