Computer Sciences Commons^™

Feedback Algorithm For Switch Location : Analysis Of Complexity And Application To Network Design, Yuriy S. Polyakov

Theses

An accelerated feedback algorithm to solve the single-facility minisum problem is studied with application to designing networks with the star topology. The algorithm, in which the acceleration with respect to the Weiszfeld procedure is achieved by multiplying the current Weiszfeld iterate by an accelerating feedback factor, is shown to converge faster than the accelerating procedures available in the literature. Singularities encountered in the algorithm are discussed in detail. A simple practical exception handling subroutine is developed. Several applications of the algorithm to designing computer networks with the star topology are demonstrated. Applications of the algorithm as a subroutine for multi-switch …

Uber-Claws : Unsupervised Pattern Classification For Multi-Unit Extracellular Neuronal Burst Extraction, Rama Natarajan

Theses

To further an understanding of how a neuronal population generates patterns of rhythmic activity, the temporal dynamics of the group of neurons must be formalized. Essential to this pursuit, is the ability to reliably detect and separate the classes of single-unit neuronal activity from multi-unit extracellular signals recorded in a single channel. This study proposes a unified approach to automatically detect and classify single-unit bursts, and to observe the precise onset and offset of burst activity. Existing approaches to the problem fundamentally depend on the statistics of spike waveform variability, both extrinsic and intrinsic to the neuron. In contrast, the …

Towards Digital Library Service Integration, Prateek Shrivastava

Theses

Digital Library Service Integration (DLSI) aims to provide a systematic approach in integrating the services and collections of National Science and Digital Library. The National Science and Digital Library collections can share the services among themselves in a totally integrated environinent. Collections as such will require no change to plug into the DLSI architecture. Collections will keep using the services of NSDL in the similar manner as before. These services will in turn pass few parameters to the services of DLSI. With the help of these parameters, wrappers will fetch the details and priority of the users. These wrappers will …

Handgrip Pattern Recognition, Zong Chen

Dissertations

There are numerous tragic gun deaths each year. Making handguns safer by personalizing them could prevent most such tragedies. Personalized handguns, also called "smart" guns, are handguns that can only be fired by the authorized user. Handgrip pattern recognition holds great promise in the development of the smart gun.

Two algorithms, static analysis algorithm and dynamic analysis algorithm, were developed to find the patterns of a person about how to grasp a handgun. The static analysis algorithm measured 160 subjects' fingertip placements on the replica gun handle. The cluster analysis and discriminant analysis were applied to these fingertip placements, and …

A Method For Developing In-Silico Protein Homologs, Susan Mcclatchy

Theses

Computational methods for identifying and screening the most promising drug receptor candidates in the human genome are of great interest to drug discovery researchers. Successful methods will accurately identify and narrow the field of potential drug receptor candidates. This study details one such method.

The method described here begins with the assumption that novel drug receptors have high sequence similarity to established drug receptors. The similarity search program FASTA3 aligns translated sequences of the human genome to known drug receptor sequences and ranks these alignments by measuring their statistical significance. Query results returned by FASTA3 are assembled into "in-silico proteins" …

Functional Annotation And Dendrogram Representation Of Gene Expression Clustering Results, Antoneta Petkova Vladimirova

Theses

The advances in genomic sciences have created vast amounts of gene expression data. To make sense of the expression information, various techniques have been applied. Clustering is among the unsupervised methods used to group the results according to gene expression level. Dendrogram visualization allows graphical representation of the clustering. The aim of this thesis is to enhance these techniques by adding another layer of functionality, namely, annotating the dendrogram with gene functional information. Presented is an application which visualizes yeast clustering results as a dendrogram along with color-coded gene keyword annotations. Gene keyword information was extracted from a major biological …

Analysis Of Gene Expression Data Using Expressionist 3.1 And Genespring 4.2, Indu Shrivastava

Theses

The purpose of this study was to determine the differences in the gene expression analysis methods of two data mining tools, Expressionistic^TM 3.1 and GeneSpring^TM 4.2 with focus on basic statistical analysis and clustering algorithms. The data for this analysis was derived from the hybridization of Rattus norvegicus RNA to the Affymetrix RG34A GeneChip. This analysis was derived from experiments designed to identify changes in gene expression patterns that were induced in vivo by an experimental treatment.

The tools were found to be comparable with respect to the list of statistically significant genes that were up-regulated by more …

Knowledge Discovery And Modeling In Genomic Databases, Michael M. Yin

Dissertations

This dissertation research is targeted toward developing effective and accurate methods for identifying gene structures in the genomes of high eukaryotes, such as vertebrate organisms. Several effective hidden Markov models (HMMs) are developed to represent the consensus and degeneracy features of the functional sites including protein-translation start sites, mRNA splicing junction donor and acceptor sites in vertebrate genes. The HMM system based on the developed models is fully trained using an expectation maximization (EM) algorithm and the system performance is evaluated using a 10-way cross-validation method. Experimental results show that the proposed HMM system achieves high sensitivity and specificity in …

Statistical Image Analysis Of Spotted Arrays, Filippo Posta

Theses

There is a lot of systematic and specific variability in microarray experiments, this variability affects measured gene expression levels, leading to unreliable gene profiling or an heavy load of extra experiment to statistically confirm the data observed in one experiment.

The aim of this work is to systematically analyze, using statistics, the image derived from a cDNA microarray experiment to have a better understanding of this variability and thus a better confidence over the data obtained from an experiment.

Using technologies available at the Center for Applied Genomics, Newark, New Jersey. Selected images derived from different type of microarray experiments …

An Algorithm For Estimating The Quality Of Microarrays, Ajeet S. Sodhi

Theses

Microarray technology is currently one of the most valuable gene expression tools in molecular biology allowing the experimenter to simultaneously quantify the expression of thousands of genes. It is also one of the most difficult tools to use accurately as each microarray produces a large amount of information that needs to be inspected and normalized before analysis. As the size of a microarray or number of replicates increase, the use of manual inspection becomes impractical. The aim of this thesis is to introduce an algorithm that evaluates each feature of a microarray from the scanned data file. A quality score …

Intramural, Collaborative Learning Systems, Robert S. Friedman

Theses

This thesis focuses on three related concepts: problem-based collaborative learning; the use of multimedia tools in learning systems; and participatory design as a software engineering methodology to create multimedia tools to be used in learning systems. A literature review of the three areas is followed by an overview of the pedagogical, technological, and business trends that affect the direction of innovation in education, including problem-based learning. A discussion of a software engineering project to develop a multimedia application that enhances the learning of geography skills and puts the programming, interface design and multimedia systems capabilities of college students into action …

Development And Characterization Of Methodology And Technology For The Alignment Of Fmri Time Series, Carlo Ciulla

Dissertations

This dissertation has developed, implemented and tested a novel computer based system (AUTOALIGN) that incorporates an algorithm for the alignment of functional Magnetic Resonance Image (fMRI) time series. The algorithm assumes the human brain to be a rigid body and computes a head coordinate system on the basis of three reference points that lie on the directions correspondent to two of the eigenvectors of inertia of the volume, at the intersections with the head boundary. The eigenvectors are found weighting the inertia components with the voxel's intensity values assumed as mass. The three reference points are found in the same …

Collaborative Problem Solving And Program Development Model, Joanna Defranco-Tommarello

Dissertations

A model to enhance collaborative problem solving and program development is presented. The collaborative model is a detailed cognitive model that takes into consideration the cognitive and social activities that occur during collaborative problem solving and program development. The cognitive activities required for collaborative problem solving and program development are identified and integrated into a six-stage model. An extensive literature review in the associated fields is presented to show the need for the model described in this dissertation. In addition, a comprehensive study of tools to support collaboration during problem solving and program development was also performed as well as …

An Evolving Approach To Learning In Problem Solving And Program Development : The Distributed Learning Model, Idania Espinosa

Theses

Technological advances are paving the way for improvements in many sectors of society. The US education system needs to undergo a transformation of existing pedagogical methods to maximize utilization of new technologies. Traditional education has primarily been teacher driven, lectured-based in one location. Advances in technology are challenging existing paradigms by developing tools and educational environments that reach diverse learning styles and surpass the boundaries of current teaching methods.

Distributed learning is an emerging paradigm today that has promise to contribute significantly to learning and improve overall academic success. This research first explores various systems that provide different modes of …

Comparative Molecular Field Analysis (Comfa) Of Protonated Methylphenidate Phenyl-Substituted Analogs, Kathleen Mary Gilbert

Theses

Protonated methylphenidate (pMP) and several phenyl-substituted pMP analogs were analyzed using Comparative Molecular Field Analysis (CoMFA) to develop a pharmacophore for dopamine transporter (DAT) binding. This research is a part of an interdisciplinary study on using methylphenidate (MP) analogs to block the binding of cocaine to the DAT as a treatment for addiction.

A random search conformational analysis using key pMP torsional angles was performed to create conformer families representing possible bioactive conformations. The lowest energy pMP conformer of each family was used as a template to create phenyl-substituted pMP analogs.

Partial least squares analysis was used to determine the …

Simulation And Close-To-Optimal Algorithm For The Static Load Balancing Of A Network Of Heterogeneous Processors, Mark G. Steiner

Theses

A close-to-optimal linear programming-based algorithm for the static load balancing of a network of heterogeneous processors is described and implemented. Experimental results suggest that the amount of time required by the implementation of the algorithm to balance the loads of the servers as a function of the number of servers has polynomial complexity.

A Review And Assessment Of Novice Learning Tools For Problem Solving And Program Development, Christopher John Bladek

Theses

There is a great demand for the development of novice learning tools to supplement classroom instruction in the areas of problem solving and program development. Research in the area of pedagogy, the psychology of programming, human-computer interaction, and cognition have provided valuable input to the development of new methodologies, paradigms, programming languages, and novice learning tools to answer this demand.

Based on the cognitive needs of novices, it is possible to postulate a set of characteristics that should comprise the components an effective novice-learning tool. This thesis will discover these characteristics and provide recommendations for the development of new learning …

A Tool For Phylogenetic Data Cleaning And Searching, Viswanath Neelavalli

Theses

Data collection and cleaning is a very important part of an elaborate Data Mining System. 'TreeBASE' is a relational database of phylogenetic information at the Harvard University with a keyword based searching interface. 'TreeSearch' is a Structure based search engine implemented at NJIT that can be used for searching phylogenetic data. Phylogenetic trees are extracted from the flat-file database at Harvard University, available at {ftp://herbaria.harvard.edu/pub/piel/Data/files/}. There is huge amount of information present in the files about the trees and the data matrices from which the trees are generated. The search tool implemented at NJIT is interested in using the string …

Computationally Efficient Search For Large Primes, Wieslawa E. Amber

Theses

To satisfy the speed of communication and to meet the demand for the continuously larger prime numbers, the primality testing and prime numbers generating algorithms require continuous advancement. To find the most efficient algorithm, a need for a survey of methods arises. Concurrently, an urge for the analysis of algorithms' performances emanates. The critical criteria in the analysis of the prime numbers generation are the number of probes, number of generated primes, and an average time required in producing one prime. Hence, the purpose of this thesis is to indicate the best performing algorithm. The survey the methods, establishment of …

E-Docspros : Exploring Texpros Into E-Business Era, Zhenfu Cheng

Dissertations

Document processing is a critical element of office automation. TEXPROS (TEXt PROcessing System) is a knowledge-based system designed to manage personal documents. However, as the Internet and e-Business changed the way offices operate, there is a need to re-envision document processing, storage, retrieval, and sharing. In the current environment, people must be able to access documents remotely and to share those documents with others. e-DOCPROS (e-DOCument PROcessing System) is a new document processing system that takes advantage of many of TEXPROS's structures but adapts the system to this new environment. The new system is built to serve e-businesses, takes advantage …

Waqs : A Web-Based Approximate Query System, George Jyh-Shian Chang

Dissertations

The Web is often viewed as a gigantic database holding vast stores of information and provides ubiquitous accessibility to end-users. Since its inception, the Internet has experienced explosive growth both in the number of users and the amount of content available on it. However, searching for information on the Web has become increasingly difficult. Although query languages have long been part of database management systems, the standard query language being the Structural Query Language is not suitable for the Web content retrieval.

In this dissertation, a new technique for document retrieval on the Web is presented. This technique is designed …

Augmenting Applications With Hyper Media, Functionality And Meta-Information, Roberto Galnares

Dissertations

The Dynamic Hypermedia Engine (DHE) enhances analytical applications by adding relationships, semantics and other metadata to the application's output and user interface. DHE also provides additional hypermedia navigational, structural and annotation functionality. These features allow application developers and users to add guided tours, personal links and sharable annotations, among other features, into applications. DHE runs as a middleware between the application user interface and its business logic and processes, in a n-tier architecture, supporting the extra functionalities without altering the original systems by means of application wrappers.

DHE automatically generates links at run-time for each of those elements having relationships …

Knowledge-Based Document Retrieval With Application To Texpros, Fang Sheng

Dissertations

Document retrieval in an information system is most often accomplished through keyword search. The common technique behind keyword search is indexing. The major drawback of such a search technique is its lack of effectiveness and accuracy. It is very common in a typical keyword search over the Internet to identify hundreds or even thousands of records as the potentially desired records. However, often few of them are relevant to users' interests.

This dissertation presents knowledge-based document retrieval architecture with application to TEXPROS. The architecture is based on a dual document model that consists of a document type hierarchy and, a …

Decision Making Based On Quantitative And Qualitative Evaluations, Rohan A. Pandit

Theses

This study emphasizes mainly on the influence of evaluations, both qualitative and quantitative, on decision making for many occasions that occur in business and technically oriented settings. Decisions made with a certain fuzzy as wen as technical behavior are structured by means of computer-assisted decision-making tools. Decision support tools assist decision makers in making crucial decisions. For instance the tool that has been designed for the purpose of this research will be used for selecting capital-intensive products. It is also intended to prove that with the help of decision support systems decision makers could make decisions by reducing fuzzy decision …

Data Mining Using Neural Networks For Large Credit Card Record Sites, Wei Wei

Theses

Data mining using neural networks has been applied in various financial fields such as risk mitigation, missing data filling, fraud detection, and customer profile classification etc. This master thesis work aims to develop methodologies to mine large sets of records and in particular to fill missing data in these records. The steps include data cleansing, data selection, data preprocessing, data representation, data clustering and finally the missing data filling. Furthermore, this work designs algorithms to evaluate the supervised neural networks' performance, which is helpful for the future research on data prediction and classification. A case study based on a large …

A More Efficient Document Retrieval Method For Texpros, Yin Dong

Dissertations

Document processing is a critical element of office automation. Through document classification, extraction and filing, documents are automatically placed into a knowledge base according to certain rules. Document retrieval is a process to get a document back according to a user's requirements and to show the results to the user. Hence, a good user-interface and an efficient retrieval algorithm become core parts of document retrieval.

Unlike previous browsers that have been proposed for this purpose, this dissertation develops a new browser that has a user interface with more tools, and one that has a more efficient retrieval algorithm that can …

Visualization Techniques For Routing Protocols And Router Configurations, Vandana Pursnani

Theses

An autonomous system (AS) is a group of routers managed by a particular organization. Exterior gateway protocols (EGP) are used between AS's Internal Gateway Protocols (IGP) is used within an AS. The most common protocols used with TCP/IP are RIP, OSPF (Open Shortest Path First), IGRP / Enhanced IGRP.

The thesis revolves around OSPF protocol OSPF uses flooding to exchange link-state updates between routers. Any change in routing information is flooded to all routers in the network. Areas are introduced to put a boundary on the explosion of link-state updates. Flooding and calculation of the Dijkstra algorithm on a router …

Towards Hypermedia Support In Database Systems, Anirban Bhaumik

Theses

The general goal of our research is to automatically generate links and other hypermedia related services to analytical applications. Using a dynamic hypermedia engine (DHE), the following features have been automated for database systems. Based on the database's relational (physical) schema and its original (non-normalized) entity-relationship specification links are generated, database application developers may also specify the relationship between different classes of database elements. These elements can be controlled by the same or different database application, or even by another software system. A DHE prototype has been developed and illustrates the above for a relational database management system.

The DHE …

Managing Toxic And Hazardous Substances Of Concern In Manufacturing, Craig Jancerak

Theses

This thesis examined data management from an object relational database perspective. The database used was a product from Cornell University called PREDATOR. The context of the evaluation is the life-cycle of a manufactured product. The manufacturing life-cycle was chosen to correspond to an environmental life-cycle being done at NJIT. However, the goal was to use a generic model that could be applied to many situations.

The first phase developed the conceptual life-cycle model. At a high level, the model was not designed with a programming language in mind. Several questions needed to be answered and the a [sic] model answers …

Information Retrieval And Mining In High Dimensional Databases, Xiong Wang

Dissertations

This dissertation is composed of two parts. In the first part, we present a framework for finding information (more precisely, active patterns) in three dimensional (3D) graphs. Each node in a graph is an undecoraposable or atomic unit and has a label. Edges are links between the atomic units. Patterns are rigid substructures that may occur in a graph after allowing for an arbitrary number of whole-structure rotations and translations as well as a small number (specified by the user) of edit operations in the patterns or in the graph. (When a pattern appears in a graph only after the …