An automated detection process for Eukaryotic ribonucleic acid (RNA) Polymerase II Promoter is presented in this dissertation. We employ an artificial neural network (ANN) in conjunction with features that were selected using an information-theoretic approach. Firstly an introduction is given where the problem is described briefly. Some background is given about the biological and genetic principles involved in DNA, RNA and Promoter detection. The automation process is described with each step given in detail. This includes the data information gathering, feature generation, and the full ANN process. The ANN section of the project is split up in a generation process, a training section as well as a testing section. Lastly the final detection program was tested and compared to other promoter detection systems. An improvement of at least 10% in positive prediction value (PPV) in comparison with current state-of-the-art solutions was obtained. Note: A Companion CD should accompany this report that contains all the program code and some of the source data that was used in this project. All the references to “Companion CD”, reference number [18] are references to these programs.acquisition process, how the different samples were split into different sets and statistical. / Dissertation (MEng (Computer Engineering))--University of Pretoria, 2007. / Electrical, Electronic and Computer Engineering / unrestricted

Increased production and use of multimedia data has led to the development of a more advanced Database Management System (DBMS), like an Object Relational Database Management System (ORDBMS). These advanced databases are necessitated by the complexity in structure and the functionality required by multimedia data. Unfortunately, no suitable benchmarks exist with which to test the performance of databases when handling multimedia data. This thesis describes the design of a benchmark to measure the performance of basic functionality found in multimedia databases. The benchmark, called MORD (Multimedia Object Relational Databases), targets Oracle, a well known commercial Object Relational Database Management System (ORDBMS) that can handle multimedia data. Although MORD targets Oracle, it can easily be applied to other Multimedia Database Management System (MMDBMS) as a result of a design that stressed its portability, and simplicity. MORD consists of a database schema, test data, and code to simulate representative queries on multimedia databases. A number of experiments are described that validate MORD and ensure its correct design and that its objectives are met. A by-product of these experiments is an initial understanding of the performance of multimedia databases. The experiments show that with multimedia data the buffer cache should be at least large enough to hold the largest dataset, a bigger block size improves the performance, and turning off logging and caching for bulk loading improves the performance. MORD can be used to compare different ORDBMS or to assist in the configuration of a specific database.

Ce mémoire présente la conception de différents systèmes antennaires miniatures devant fonctionner dans des conditions environnementales et climatiques sévères, mais également capables d’opérer dans plusieurs bandes de fréquence ISM et plusieurs standards de communications mobiles. Ce système antennaire dédié infine à l’intégration dans les modules TRAX-BOX développés par la société TRAXENS, doit satisfaire différents critères techniques, industriels et économiques, définis par leur application finale qui est le suivi de conteneurs métalliques multimodaux de grandes dimensions. Pour cela, le système doit présenter un fonctionnement multibande et large bande, un diagramme de rayonnement omnidirectionnel, une bonne efficacité, un encombrement réduit permettant une intégration simple et efficace dans son boîtier, une rigidité mécanique permettant au système d’être protégé contre les vibrations et les chocs, une large plage de températures de fonctionnement et un faible coût de fabrication. Ces exigences techniques et industrielles sont souvent difficiles à satisfaire simultanément. Pour ce faire, plusieurs techniques de conception de systèmes multi-antennaires multibandes ont été employées. L’insertion de fentes sur les résonateurs et la juxtaposition d’éléments parasites a tout d’abord permis d’élargir la bande des antennes. En ce qui concerne l’isolation mutuelle des antennes, l’insertion de fentes sur le plan de masse, est l’une des techniques qui permettent d’isoler les antennes. / This thesis presents the design of different small antenna systems, operating in harsh environmental and climatic conditions, but also be able to operate in several ISM frequency bands and several mobile communication standards. This antenna system designed to be integrated in the TRAXBOX modules developed by TRAXENS, must meet various technical, industrial and economic criteria, defined by their final application which is the monitoring of multimodal metallic containers. To this end, the system must provide a multiband and wideband operations, an omnidirectional radiation pattern, a good efficiency, a reduced size allowing an easy and efficient integration into its housing, mechanical rigidity allowing the system to be protected against vibrations and shocks, a wide operating temperature range and low manufacturing cost. These technical and manufacturing requirements are often difficult to simultaneously satisfy. To do this, several design techniques of multiband multi-antenna systems have been employed. The use of slots and parasitic elements has first allowed widening the antennas bands. Regarding the mutual isolation of the antennas, the insertion of slots on the ground plane is one of the techniques that improve the antennas isolation. However, in this manuscript, we have mainly focused on the optimal layout of the parasitic resonators to simultaneously increase the bandwidth but also optimize the system efficiency and the isolation between antennas after interpretation of the various physical phenomena observed.

Poor management of nitrogen (N) fertilisers and water in agro-ecosystems reduces yield, quality and N-use efficiency, and leads to pollution. The objective of this study was to improve irrigation and N management for planted pastures through adaptive management with simple tools and modelling. Field experiments were conducted in 2007 and 2008 at Cedara (KwaZulu Natal) and Hatfield (Gauteng) using annual ryegrass as a case study under a range of N and irrigation application strategies. Collected data sets were also used to calibrate and validate the SWB-Pro (simple) and SWB-Sci (detailed) model versions. After validation, the model was used to develop irrigation calendars and strategies, and estimate irrigation requirements for annual ryegrass. The highest forage yields were produced when N application rates ranged between 30 to 60 kg N ha-1 for each growth cycle, except for the first 2-3 growth cycles when there was high soil N carryover from the previous season. The current farmers’ recommendation (fixed N application rate of 50 kg ha-1 per growth cycle) maximised biomass but reduced pasture quality. Adaptive strategies based on nitrate concentration in wetting front detectors at different depths, reduced fertiliser N application by 28–32% and reduced potentially leachable residual soil N, while improving forage quality without yield reduction. The rate 30-40 kg N ha-1 per growth cycle provided a compromise between forage yield and quality. The SWB model performed well in simulating ryegrass growth, leaf area index, forage yield, root zone soil water deficit, daily evapotranspiration, biomass N uptake and soil nitrate. Site specific and monthly variable irrigation calendars were developed using the SWB-Pro model, for four major milk producing areas of South Africa. The simpler monthly irrigation calendars can be used in the absence of irrigation monitoring tools or more accurate site specific calendars. The SWB-Pro model requires relatively few and simple inputs. However, irrigation monitoring/scheduling with the aid of real time modelling or measurements is better than calendars developed using the SWB-Pro model with long-term historical weather data. The SWB-Sci model showed ways of improving water use efficiency using ‘room for rain’ and ‘mildly deficit irrigation’ approaches in high rainfall areas. Scenario modelling demonstrated that the best management strategy of achieving maximum yield together with low N leaching is by integrating N and water management. This integrated management can be based on the wetness of the soil and nitrate concentration in the deep root zone using wetting front detectors. The model can be used to generate monitoring protocols such as depth of wetting front detector placement and selecting N thresholds to be used for adaptive management. Setting approximate thresholds for wetting depth and nitrate concentration is a first step in implementing an adaptive management strategy. However, the challenge is to find monitoring tools which allow effective implementation of the strategy. In this study, the wetting front detector proved to be a robust, on-farm water and nitrate monitoring tool which is relatively simple and cost effective. Should it become widely adopted, farmers are expected to improve these thresholds as more experience is gained. The SWB model could also be used to evaluate alternative thresholds for adaptive N and water management. / Thesis (PhD)--University of Pretoria, 2012. / Plant Production and Soil Science / unrestricted

The effectiveness of implementing field programmable gate arrays (FPGAs) in communication, military, space and high radiation environment applications, coupled with the increased accessibility of private individuals and researchers to launch satellites, has led to an increased interest in commercial off the shelf components. The metal oxide semiconductor (MOS) structures of FPGAs however, are sensitive to radiation effects which can lead to decreased reliability of the device. In order to successfully implement a FPGA based system in a radiation environment, such as on-board a satellite, the single event upset (SEU) and total ionizing dose (TID) characteristics of the device must first be established. This research experimentally determines a research procedure which could accurately determine the SEU cross sections and TID characteristics of various mitigation techniques as well as control circuits implemented in a ProASIC3 A3P1000 FPGA. To gain an understanding of the SEU effects of the implemented circuits, the test FPGA was irradiated by a 66MeV proton beam at the iTemba LABS facility. Through means of irradiation, the SEU cross section of various communication, motor control and mitigation schemes circuits, induced by high energy proton strikes was investigated. The implementation of a full global triple modular redundancy (TMR) and a combination of TMR and a AND-OR multiplexer filter was found to most effectively mitigate SEUs in comparison to the other techniques. When comparing the communication and motor control circuits, the high frequency I2C and SPI circuits experienced a higher number of upsets when compared to a low frequency servo motor control circuit. To gain a better understanding of the absorbed dose effects, experimental TID testing was conducted by irradiating the test FPGA with a cobalt-60 (Co-60) source. An accumulated absorbed dose resulted in the fluctuation of the device supply current and operating voltages as well as resulted in output errors. The TMR and TMR filtering combination mitigation techniques again were found to be the most effective methods of mitigation.

Penetration testing is widely used in industry as a test method for web application security assessment. However, penetration testing is often performed late in a software development life cycle as an isolated task and usually requires specialized security experts. There is no well-defined test framework providing guidance and support to general testers who usually do not have in-depth security expertise to perform a systematic and cost-efficient penetration test campaign throughout a security-oriented software development life cycle. In this thesis, we propose a model-driven penetration test framework for web applications that consists of a penetration test methodology, a grey-box test architecture, a web security knowledge base, a test campaign model, and a knowledge-based PenTest workbench. The test framework enables general testers to perform a penetration test campaign in a model-driven approach that is fully integrated into a security-oriented software development life cycle. Security experts are still required to build up and maintain a web security knowledgebase for test campaigns, but the general testers are capable of developing and executing penetration test campaigns with reduced complexity and increased reusability in a systematic and cost-efficient approach. A prototype of the framework has been implemented and applied to three web applications: the benchmark WebGoat web application, a hospital adverse event management system (AEMS), and a palliative pain and symptom management system (PAL-IS). An evaluation of the test framework prototype based on the case studies indicates the potential of the proposed test framework to improve how penetration test campaigns are performed and integrated into a security-oriented software development life cycle.

Web applications have come a long way both in terms of adoption to provide information and services and in terms of the technologies to develop them. With the emergence of richer and more advanced technologies such as Ajax, web applications have become more interactive, responsive and user friendly. These applications, often called Rich Internet Applications (RIAs) changed the traditional web applications in two primary ways: Dynamic manipulation of client side state and Asynchronous communication with the server. At the same time, such techniques also introduce new challenges. Among these challenges, an important one is the difficulty of automatically crawling these new applications. Crawling is not only important for indexing the contents but also critical to web application assessment such as testing for security vulnerabilities or accessibility. Traditional crawlers are no longer sufficient for these newer technologies and crawling in RIAs is either inexistent or far from perfect. There is a need for an efficient crawler for web applications developed using these new technologies. Further, as more and more enterprises use these new technologies to provide their services, the requirement for a better crawler becomes inevitable. This thesis studies the problems associated with crawling RIAs. Crawling RIAs is fundamentally more difficult than crawling traditional multi-page web applications. The thesis also presents an efficient RIA crawling strategy and compares it with existing methods.

With more and more images being uploaded to social networks each day, the resources for identifying a large portion of the world are available. However the tools to harness and utilize this information are not sufficient. This thesis presents a system, called PhacePhinder, which can build a face database from a social network and have it accessible from mobile devices. Through combining existing technologies, this is made possible. It also makes use of a fusion probabilistic latent semantic analysis to determine strong connections between users and content. Using this information we can determine the most meaningful social connection to a recognized person, allowing us to inform the user of how they know the person being recognized. We conduct a series of offline and user tests to verify our results and compare them to existing algorithms. We show, that through combining a user’s friendship information as well as picture occurrence information, we can make stronger recommendations than based on friendship alone. We demonstrate a working prototype that can identify a face from a picture taken from a mobile phone, using a database derived from images gathered directly from a social network, and return a meaningful social connection to the recognized face.

In this work, we provide a qualitative comparison between existing geocast protocols and then we present an efficient geocast routing protocol for VANET. This protocol is a semantic and self-decision geocast routing protocol for disseminating safety and non-safety information over VANET (SAS-GP). SAS-PG initially executes an algorithm to locally determine the semantic geocast area. Then, the protocol disseminates the information in three phases: Spread, Preserve, and Assurance, which utilize the traffic information system and the digital map. SAS-GP principally employs timer-based techniques in order to avoid overhead and broadcast storm problems; nonetheless, novel factors are enhanced to calculate the timer’s values in each phase. Simulation results demonstrate effective and reliable dissemination in terms of delivery ratio and number of false warnings compared to existing protocols when evaluated in high scale and realistic scenarios. Also, SAS-GP performs faster in notifying vehicles resulting in a higher geocast distance before approaching the location of the event.

As important components of intelligent transportation system, lane detection and tracking (LDT) and lane departure warning (LDW) systems have attracted great interest from the computer vision community over the past few years. Conversely, lane markings recognition (LMR) systems received surprisingly little attention. This thesis proposed a real-time lane assisting framework for intelligent vehicles, which consists of a comprehensive module and simplified module. To the best of our knowledge, this is the first parallel architecture that considers not only lane detection and tracking, but also lane marking recognition and departure warning. A lightweight version of the Hough transform, PPHT is used for both modules to detect lines. After detection stage, for the comprehensive module, a novel refinement scheme consisting of angle threshold and segment linking (ATSL) and trapezoidal refinement method (TRM) takes shape and texture information into account, which significantly improves the LDT performance. Also based on TRM, colour and edge informations are used to recognize lane marking colors (white and yellow) and shapes (solid and dashed). In the simplified module, refined MSER blobs dramatically simplifies the preprocessing and refinement stage, and enables the simplified module performs well on lane detection and tracking. Several experiments are conducted in highway and urban roads in Ottawa. The detection rate of the LDT system in comprehensive module average 95.9% and exceed 89.3% in poor conditions, while the recognition rate depends on the quality of lane paint and achieves an average accuracy of 93.1%. The simplified module has an average detection rate of 92.7% and exceeds 84.9% in poor conditions. Except the conventional experimental methods, a novel point cluster evaluation and pdf analysis method have been proposed to evaluate the performance of LDT systems, in terms of the stability, accuracy and similarity to Gaussian distribution.