No description available.

This manuscript consists of three papers which formulate novel generalized linear model methodologies. In Chapter 1, we introduce a variant of the traditional concordance statistic that is associated with logistic regression. This adjusted c − statistic as we call it utilizes the differences in predicted probabilities as weights for each event/non- event observation pair. We highlight an extensive comparison of the adjusted and traditional c-statistics using simulations and apply these measures in a modeling application. In Chapter 2, we feature the development and investigation of three model selection criteria based on cross-validatory c-statistics: Model Misspecification Pre- diction Error, Fitting Sample Prediction Error, and Sum of Prediction Errors. We examine the properties of the corresponding selection criteria based on the cross- validatory analogues of the traditional and adjusted c-statistics via simulation and illustrate these criteria in a modeling application. In Chapter 3, we propose and investigate an alternate approach to pseudo- likelihood model selection in the generalized linear mixed model framework. After outlining the problem with the pseudo-likelihood model selection criteria found using the natural approach to generalized linear mixed modeling, we feature an alternate approach, implemented using a SAS macro, that obtains and applies the pseudo-data from the full model for fitting all candidate models. We justify the propriety of the resulting pseudo-likelihood selection criteria using simulations and implement this new method in a modeling application.

In a Biostatistics environment, the datasets to be analysed are frequently high-dimensional and multicollinearity is expected due to the nature of the features. However, many traditional approaches to statistical analysis and feature selection cease to be useful in the presence of high-dimensionality and multicollinearity. Penalised regression methods have proved to be practical and attractive for dealing with these problems. In this dissertation, we propose a new penalised approach, the modified elastic-net (MEnet), for statistical analysis and feature selection using a combination of the ridge and bridge penalties. This method is designed to deal with high-dimensional problems with highly correlated predictor variables. Furthermore, it has a closed-form solution, unlike the most frequently used penalised techniques, which makes it simple to implement on high-dimensional data. We show how this approach can be used to analyse high-dimensional data with binary responses, e.g., microarray data, and simultaneously select significant features. An extensive simulation study and analysis of a colon cancer dataset demonstrate the properties and practical aspects of the proposed method. / Mini Dissertation (MSc (Advanced Data Analytics))--University of Pretoria, 2020. / DSI-CSIR Interbursary Support (IBS) Programme / Statistics Industry HUB, Department of Statistics, University of Pretoria / Statistics / MSc / Restricted

Proteins play critical roles in cellular processes of living organisms. It is therefore important to identify and characterize their key properties associated with their functions. Correlating protein’s structural, sequence and physicochemical properties of its amino acids (aa) with protein functions could identify some of the critical factors governing the specific functionality. We point out that not all functions of even well studied proteins are known. This, complemented by the huge increase in the number of newly discovered and predicted proteins, makes challenging the experimental characterization of the whole spectrum of possible protein functions for all proteins of interest. Consequently, the use of computational methods has become more attractive. Here we address two questions. The first one is how to use protein aa sequence and physicochemical properties to characterize a family of proteins. The second one focuses on how to use transcription factor (TF) protein’s domains to enhance accuracy of predicting TF DNA binding sites (TFBSs). To address the first question, we developed a novel method using computational representation of proteins based on characteristics of different protein regions (N-terminal, M-region and C-terminal) and combined these with the properties of protein aa sequences. We show that this description provides important biological insight about characterization of the protein functional groups. Using feature selection techniques, we identified key properties of proteins that allow for very accurate characterization of different protein families. We demonstrated efficiency of our method in application to a number of antimicrobial peptide families. To address the second question we developed another novel method that uses a combination of aa properties of DNA binding domains of TFs and their TFBS properties to develop machine learning models for predicting TFBSs. Feature selection is used to identify the most relevant characteristics of the aa for such modeling. In addition to reducing the number of required models to only 14 for several hundred TFs, the final prediction accuracy of our models appears dramatically better than with other methods. Overall, we show how to efficiently utilize properties of proteins in deriving more accurate solutions for two important problems of computational biology and bioinformatics.

The purpose of this thesis is to set forth the ideals and the philosophy of reading for young people and to show the importance of' stimulating and guiding students in all phases of their reading so that they may find increasing enjoyment and satisfaction and may grow in critical judgment and appreciation. It is therefore necessary that the books from which they choose their reading are of the best available, and so it is important that the librarian give the greatest thought and study to the selection of books for her library. To make these ideals practical and usable, carefully prepared lists of books have been made for various purposes. These lists are constantly used by both teachers and pupils at Stanford, and have been a source of help and inspiration. There are many excellent books on children's reading and the selection of books for children of different ages in different fields of interest. It has been necessary to limit the field of research and discussion and it has not been attempted to make an exhaustive list or selection for all material in the field of book selection. The thesis can only be suggestive, as it is a field which grows and one in which new material is constatnly appearing.

The main aim of this dissertation is to address the security issues of the physical layer of Cyber Physical Systems. The network security is first assessed using a 5-level Network Security Evaluation Scheme (NSES). The network security is then enhanced using a novel Intrusion Detection System that is designed using Supervised Machine Learning. Defined as a complete architecture, this framework includes a complete packet analysis of radio traffic of Routing Protocol for Low-Power and Lossy Networks (RPL). A dataset of 300 different simulations of RPL network is defined for normal traffic, hello flood attack, DIS attack, increased version attack and decreased rank attack. The IDS is a multi-model detection model that provides an efficient detection against the known as well as new attacks. The model analysis is done with the cross-validation method as well as using the new data from a similar network. To detect the known attacks, the model performed at 99% accuracy rate and for the new attack, 85% accuracy is achieved. / Graduate

The purpose of this study was to examine how recruitment and selection processes are conducted at the Robben Island Museum and the challenges associated with it. Correct implementation of the recruitment and selection practices is crucial in order for the organisation to fill the right positions with the right people who are experienced and competent. In other words, organisations should strive for excellence in ensuring that there is conformity to legal prescripts whenever the recruitment and selection of employees commence. A qualitative research design was applied in order to achieve the primary aim of this study. Data collection techniques that were utilised to collect information comprised interviews and document analysis. A group of fourteen (14) purposively selected participants, namely seven operational staff members and seven managers were chosen for interviews. Data that was obtained was analysed through qualitative content analysis. The major findings of the study indicate that the Robben Island Museum recruits potential candidates through various means such as newspaper advertisements, employment agencies, headhunting, job posting and online recruitment. The study found that although qualifications are seen to be necessary during the recruitment and selection processes but do not seem to be seriously considered as part of the selection criteria. Essentially, the study identified some inconsistencies and failure to adhere to the recruitment and selection policy during recruitment and selection processes. Therefore, the processes of recruiting and selecting potential employment candidates should be undertaken in accordance with organisational policies and in a professional manner. / Public Administration and Management

Thesis (M.S.)--Boston University, Henry M. Goldman School of Graduate Dentistry, 1987 (Dental Public Health). / Includes bibliographical references (leaves 63-66). / The shortage of clinical patients utilizing the services provided at dental schooI clinics is a concern shared by many institutions. In an effort to help increase the patient pool at BostOn University Goldman School of Graduate Dentistry,　dental screening programs were begun in 1983, which focused at targeting　the college student population as prospective clinical patients. The following study was designed to evaluate the effectiveness of the screening programs,　conducted during the fall of 1986, in recruiting dental patients for the school and ascertain specific reasons which expediate or preclude college students from utilizing the dental services at Boston University.　The results of the study demonstrate the moderate success of the screening programs and indicate the need for continued efforts in improving the visibility and organization of the screening sessions at the college institutions participating in the Dental Screening Programs.

In this thesis, we investigate how well we can reconstruct the best rank-? approximation of a large matrix from a small number of its entries. We show that even if a data matrix is of full rank and cannot be approximated well by a low-rank matrix, its best low-rank approximations may still be reliably computed or estimated from a small number of its entries. This is especially relevant from a statistical viewpoint: the best low-rank approximations to a data matrix are often of more interest than itself because they capture the more stable and oftentimes more reproducible properties of an otherwise complicated data-generating model. In particular, we investigate two agnostic approaches: the first is based on spectral truncation; and the second is a projected gradient descent based optimization procedure. We argue that, while the first approach is intuitive and reasonably effective, the latter has far superior performance in general. We show that the error depends on how close the matrix is to being of low rank. Our results can be generalized to the spectral and entrywise error and provide flexible tools for the error analysis of the follow-up computation. Moreover, we derive a high-order decomposition of the error. With an explicit expression of the main error source, we obtain an improved estimate of the linear form. Both theoretical and numerical evidence is presented to demonstrate the effectiveness of the proposed approaches.

In affinity protein biotechnology the selective bindingbetween a chosen protein and an interacting biomolecule isutilized for a variety of applications including bioseparation,detection and therapy. Traditionally, affinity proteinsrecruited for such applications have been derived from naturalproteins or immunoglobulins generated via immunization routes.More recently, advances in the construction and handling oflarge collections of proteins(denoted libraries) generated invitro have opened up for new routes for the development ofaffinity proteins with desired properties. In this study, phage display selection technology was usedfor the isolation of novel human CD28 (hCD28)-specific affinityproteins from a protein library constructed by combinatorialprotein engineering of a 58 aa protein domain (Z) derived fromstaphylococcal protein A (SPA). From selections using hCD28 asa target molecule, several hCD28-specific affinity proteins(denoted affibodies) could be identified and analysis of theisolated affibody variants revealed a high degree of sequencehomology between the different clones. The biosensor analysisshowed that all variants bound to hCD28 with micromolardissociation constants (KD) and no significant cross-reactivitytowards the structurally related T-cell receptor hCTLA-4 couldbe observed. The apparent binding affinity for hCD28 of one ofthe isolated affibodies was further improved through fusion toa human Fc fragment fusion partner, resulting in a homodimericversion of the affibody ligand showing avidity effects uponhCD28 binding. Further, a co-culture experiment involvingJurkat T-cells and CHO cell lines tranfected to express eitherhuman CD80 or LFA-3 on the cell surface showed that apreincubation of Jurkat cells with one of the affibody variantsresulted in a specific concentration-dependent inhibition ofthe CD80 induced IL-2 production. This indicates that thisaffibody binds to hCD28 and specifically interferes with theco-stimulation signal mediated via hCD28 and hCD80. ACD28-specific binding protein could have potential as an agentfor various immunotherapy applications. In a second study, anaffinity protein-based strategy was investigated forsite-specific anchoring of proteins onto cellulose for woodfiber engineering purposes. Here, affinity proteins derivedfrom different sources were used for the assembly of acellulosome-like complex for specific and reversible anchoringof affinity domain-tagged reporter proteins to acellulose-anchored fusion protein. A fusion protein between acellulose binding module (Cel6A CBM1) derived from the fungalTrichoderma reesei and a five-domain staphylococcal protein A(SPA) moiety was constructed to serve as a platform for thedocking of reporter proteins produced as fusion to two copiesof a SPA-binding affibody affinity protein (denoted ZSPA-1),selected by phage display technology from a Z domain basedprotein library. In a series of experiments, involving repeatedwashing and low pH elutions, affinity tagged Enhanced GreenFluorescent Protein (EGFP) and Fusarium solani pisi lipasecutinase reporter proteins were both found to be specificallydirected from solution to a region of a cellulose-based filterpaper where the SPA-CBM fusion protein previously had beenpositioned. This showed that the cellulose-anchored SPA-Cel6ACBM1 fusion protein had been stably anchored to the surfacewith retained binding activity and that the interaction betweenSPA and the ZSPA-1 affibody domain was selective. phage display, combinatorial, selection, CD28, cellulosome,cellulose, affibody / NR 20140805