Flying Camera Solutions (FlyCam), within Sony Lund's startup accelerator, intends to provide drone videography to paying customers in ski resorts: a customer should be able to go about their activity as usual while a drone films them. Visual object tracking, enabling the drone to track the customer throughout the activity, is a primary obstacle in creating a viable autonomous videography service. FlyCam needs an object tracking algorithm which is accurate, robust, real-time, and requiring minimal computational overhead. We propose two innovations to aid in the selection of an appropriate tracking algorithm. Firstly, a video annotation algorithm, making use of an object detector to record the position and type of object in each frame of a video clip. Secondly, an algorithm designed to evaluate the performance of any given object tracker based on a set of performance metrics. These metrics include, among others, measures of positional accuracy, frame rate, and false positive rate. For the video annotation algorithm we implemented the state-of-the-art Mask R-CNN object detector, which achieved an average frame rate of 1.5 fps annotating video clips in up to 4K resolution. Another algorithm then played back the annotated clips to the user such that incorrect object detections could be rooted out or rectified. With little relevant annotated video available, the annotation algorithm proved useful in preparing a suite of 18 clips to be evaluated. Ten performance metrics were adapted from multi-object to single-object tracking. Nine tracking algorithms were then run on each of the 18 test video clips at varying resolutions to produce 375 tracking observations for analysis. The evaluation results revealed the optimal tracking algorithm to be Re3: a recurrent-convolutional neural network tracker which runs at respectable speeds on a consumer laptop. This is a promising result; with enough annotated data, neural networks can be retrained to improve performance. Within just a few months of operation, FlyCam could amass enough specific video data to significantly improve the neural network-based tracker.

With the rise of schedules and scheduling problems, solutions proposed in literature have expanded yet the disconnect between research and reality remains. The University of Cape Town's (UCT) Examinations Office currently produces their schedules manually with software relegated to error-checking status. While they have requested automation, this study is the first attempt to integrate optimisation techniques into the examination timetabling process. Tabu search and Nelder-Mead methodologies were tested on the UCT November 2014 examination timetabling data with tabu search proving to be more effective, capable of producing feasible solutions from randomised initial solutions. To make this research more accessible, a user-friendly app was developed which showcased the optimisation techniques in a more digestible format. The app includes data cleaning specific to UCT's data management system and was presented to the UCT Examinations Office where they expressed support for further development: in its current form, the app would be used as a secondary tool after an initial solution has been manually obtained.

Childhood growth impacts the future welfare of an individual and ultimately the nation. The importance of childhood growth monitoring with growth curves that accurately represent the growth of the population of interest cannot be overemphasised. This dissertation sought to model the growth of a cohort of South African children and compare their growth to the World Health Organisation (WHO) 2006 Child Growth Standards. Growth reference curves were derived using parametric and semi-parametric methods within the Generalised Additive Models for Location, Scale and Shape (GAMLSS) framework. Various distributions for the growth measurements were compared as well as various curve smoothing approaches for the longitudinal profiles, including cubic splines, fractional polynomials and Berkey-Reed First and Second Order models. The preferred approach was to use the Box-Cox Power Exponential (BCPE) distribution with curve smoothing by cubic splines. Non-parametric quantile regression served as a confirmation that the chosen parametric distributions were appropriate for the data. A comparison of the derived growth references to the WHO (2006) standards revealed deviations in the patterns of growth and a greater likelihood of diagnosing a child as underweight, stunted or having micro- or macrocephaly when measured against the WHO standards. The poor socioeconomic status and associated harmful exposures of the cohort were noted as potential contributing factors. A fair comparison would require a reasonably healthy and representative sample of the South African population. These findings do however call into question the appropriateness of the WHO standards for measuring the growth of South African children and bring into focus the value of developing national growth standards.

Real-time, radar-based human activity and target recognition has several applications in various fields. Examples include hand gesture recognition, border and home surveillance, pedestrian recognition for automotive safety and fall detection for assisted living. This dissertation sought to improve the speed and accuracy of a previously developed model classifying human activity and targets using radar data for outdoor surveillance purposes. An improvement in accuracy and speed of classification helps surveillance systems to provide reliable results on time. For example, the results can be used to intercept trespassers, poachers or smugglers. To achieve these objectives, radar data was collected using a C-band pulse-Doppler radar and converted to spectrograms using the Short-time Fourier transform (STFT) algorithm. Spectrograms of the following classes were utilised in classification: one human walking, two humans walking, one human running, moving vehicles, a swinging sphere and clutter/noise. A seven-layer residual network was proposed, which utilised batch normalisation (BN), global average pooling (GAP), and residual connections to achieve a classification accuracy of 92.90% and 87.72% on the validation and test data, respectively. Compared to the previously proposed model, this represented a 10% improvement in accuracy on the validation data and a 3% improvement on the test data. Applying model quantisation provided up to 3.8 times speedup in inference, with a less than 0.4% accuracy drop on both the validation and test data. The quantised model could support a range of up to 89.91 kilometres in real-time, allowing it to be used in radars that operate within this range.

The KwaZulu Natal (KZN) province is the front runner for malaria elimination in South Africa. It accounts for a small proportion of the total number of malaria cases diagnosed in the whole country in recent times. This study focused on the key localities in the province, Ndumo, and Shemula which reported the highest number of local malaria cases between the years, 2014 and 2018. The study aimed at investigating and assessing the most influential factors that drive malaria in the localities and to represent the malaria key features such as treatment, imported cases, vector spraying, and vector/human relationships. The model used in the study examines the malaria behaviour at a smaller scale as other models mainly look at larger population sizes such as district level and provinces to find the most effective strategies as we move closer to elimination. The purpose of this was to understand how malaria will change in the future if the existing strategies change. It also aimed at studying impact these changes would have on the existing cases, as to whether there will be a rise or a drop with the existing intervention coverages. This was achieved by formulating an 11 compartmental population-based, nonlinear stochastic ordinary differential equation model that will be used to simulate malaria transmission in the two localities to assess the potential impact of various policy interventions that may be used to achieve malaria elimination. It was also developed to assess the impact of policy interventions on imported infections, seasonal spraying, the effectiveness of reducing the current coverages over time, and to reach the goal of malaria elimination. Based on our analysis, we deduced that to maintain a low number of malaria cases, it would be sufficient to employ the current coverages but to reduce the number of cases, we need to consider finding ways to increase the IRS efficacy. Thus, for IRS, we conclude that, to reduce the malaria cases to its minimum (even further to 0), we need to consider increasing both the IRS coverage and its efficacy closer to 100%. With imported cases having a big impact on local cases, we concluded that we could reduce the number of local cases if we can control imported cases from other areas. Strategies such as the border clinics, screening at the border etc, would result in significant impact in the local malaria cases as we would eliminate one of the major contributors to existing malaria cases. In conclusion, we believe that increasing our efforts on the existing interventions, would result in a further decrease in the number of cases. Although one would argue that the investment is not worthwhile and that the decrease is redundant, and because of this, it is worth considering moving all those efforts towards the prevention of the more concerning variable; imported cases. In terms of local cases, we would then consider maintaining the current coverages. In that case, we should only treat those who require the treatment and spray the areas that still report cases.

Analysis of variance could be described as a statistical technique for analysing measurements depending on several kinds of effects operating simultaneously so as to decide which kinds of effects are important and to estimate the effects. Although probably not susceptible of a very precise definition, it in general consists of a body of tests of hypotheses and methods of estimation using statistics which are linear combinations of sums of squares of linear functions of the observed values. Having been developed mainly in connection with problems of agricultural experimentation, the application thereof in the South African Wool Trade seems non existent. I hope that this thesis will illustrate some of the very useful applications, especially to the extent where the rejection of all (or some) of the hypotheses under consideration is in itself as significant as the acceptance thereof would have been.

Traditional choice decisions involve selecting a single, best alternative from a larger set of potential options. In contrast, portfolio decisions involve selecting the best subset of alternatives — alternatives that together maximize some measure of value to the decision maker and are within their available resources to implement. Examples include capital investment, R&D project selection, and maintenance planning. Portfolio decisions involve a combinatorial aspect that makes them more theoretically and computationally challenging than choice problems, particularly when there are interactions between alternatives. Several portfolio decision analysis methods have been developed over the years and an increasing interest has been noted in the field of portfolio decision analysis. These methods are typically called “exact” methods, but can also be called prescriptive methods. These are generally computationally-intensive algorithms that require substantial amounts of information from the decision maker, and in return yield portfolios that are provably optimal or optimal within certain bounds. These methods have proved popular for choice decisions — for example, those based on multiattribute value or utility theory. But whereas information and computational requirements for choice problems are probably manageable for the majority of diligent decision makers, it is much less clear that this is true of portfolio decisions. That is, for portfolio decisions it may be more common that decision makers do not have the time, expertise and ability to exert the effort to assess all the information required of an exact method. Heuristics are simple, psychologically plausible rules for decision making that limit the amount of information required and the computation effort needed to turn this information into decisions. Previous work has shown that people often use heuristics when confronted with traditional choice problems in unfacilitated contexts, and that these can often return good results, in the sense of selecting alternatives that are also ranked highly by exact methods. This suggests that heuristics may also be useful for portfolio decisions. Moreover, while the lower information demands made by choice problems mean that heuristics have not generally been seen as prescriptive options, the more substantial demands made by portfolio decisions make a priori case for considering their use not just descriptively, but as tools for decision aid. Very little work exists on the use of heuristics for portfolio decision making, the subject of this thesis. Durbach et al. (2020) proposed a family of portfolio selection heuristics known collectively as add-the-best. These construct portfolios by adding, at every step, the alternative that is best in a greedy sense, with different definitions of what “best” is. This thesis extends knowledge on portfolio heuristics in three main respects. Firstly, we show that people use certain of the add-the-best heuristics when selecting portfolios without facilitation, in a context where there are interactions between alternatives. We run an experiment involving actual portfolio decision making behaviour, administered to participants who had the opportunity to choose as many alternatives as they wanted, but under the constraint of a limited budget. This experiment, parts of which were reported in Durbach et al. (2020), provides the first demonstration of the use of heuristics in portfolio selections. Secondly, we use a simulation experiment to test the performance of the heuristics in two novel environments: those involving multiple criteria, and those in which interactions between projects may be positive (the value of selecting two alternatives is more than the sum of their individual values) or negative (the opposite). This extends the results in Durbach et al. (2020), who considered only environments involving a single criterion and positive interactions between alternatives. In doing so we differentiate between heuristics that guide the selection of alternatives, called selection heuristics, and heuristics for aggregating performance across criteria, which we call scoring heuristics. We combine various selection and scoring heuristics and test their performance on a range of simulated decision problems. We found that certain portfolio heuristics continued to perform well in the presence of negative interactions and multiple criteria, and that performance depended more on the approach used to build portfolios (selection heuristics) than on the method of aggregation across criteria (scoring heuristics). We also found that in these extended conditions heuristics continued to provide outcomes that were competitive with optimal models, but that heuristics that ignored interactions led to potentially poor results. Finally, we complement behavioral and simulation experimental studies with an application of both exact methods and portfolio heuristics in a real-world portfolio decision problem involving the selection of the best subset of research proposals out of a pool of proposals submitted by researchers applying for grants from a research institution. We provide a decision support system to this institution in the form of a web-based application to assist with portfolio decisions involving interactions. The decision support system implements exact methods, namely the linear-additive portfolio value model and the robust portfolio model, as well as two portfolio heuristics found to perform well in simulations.

This story is about the art of maquis among women in Ouagadougou, Burkina Faso. The thesis frames the art of maquis as a navigational technique through which women embody their aspirational self. Here, I conceptualize the art of maquis through the notions of makeup and making up. The women I worked with used makeup framed as a concept and a practice, where making up is considered the practice through which the art of maquis is performed. Here, the tools women employ to beautify their lives are discussed in terms of technologies of visibility and behavioural techniques such as flatter [to flatter]. This monograph examines how women constantly navigate opportunities by embodying their aspirations and intersubjectivity through an ethnographic analysis of makeup and making up practices in a maquis [local pub]. To navigating precarious conditions and the materiality of the contexts, the women I worked with used makeup for pragmatic reasons, often to access aspirations in the form of socio-economic capital (making up). Experts at the art of maquis (makeup and making up), these women use their bodily capital and technologies of visibility to attract and navigate opportunities in a world where they find themselves at the margins of global capitalism. Ultimately, focusing on eye and skin makeup, this ethnography of facial and behavioural adornment showcases how people aspire to be happy through technologies of visibility and the presentation of self in everyday life. The thesis suggests that studying adornment techniques from and through the maquis provides a nuanced way of theorizing the kaleidoscopic epistemologies informing gender constructions, contemporary beauty ideals and female agency in Ouagadougou.

In this dissertation, we will introduce a new formulation of variational auto-encoders in order to generate the data we require. Our variational auto-encoder is based on data generation principles from elementary probability i.e. finding the inverse cumulative distribution function and using uniform inputs to generate samples from the distribution. Like all autoencoders, the goal is to reduce the dimensionality in the kernel and use this to describe the data features in the generation. Our formulation will use a kernel which transforms the outputs of the encoder into multi-dimensional uniformly distributed variables, which in turn will learn the cumulative distribution function (in the case of a one dimensional latent space) or the relationship of variables to copula input uniforms (in the case of a multi-dimensional latent space). The decoder will then train to learn the inverse of the encoder and this will then be used to generate data.

Attention is drawn to the confusion which surrounds the concept of nutrition status and the problem of selecting an optimum subset of variables by which nutrition status can best be assessed is defined. Using a multidisciplinary data set of some 60 variables observed on 1898 school children from four racial groups, the study aims to identify statistically, both those variables which are unrelated to nutrition status and also those which, although related, are so highly correlated that the measurement of all would be an unnecessary extravagance. It is found that, while the somatometric variables provide a reasonably good (but non-specific) estimate of nutrition status, the disciplines form meaningful groups and the variables of the various disciplines tend to supplement rather than replicate each other. Certain variables from most of the disciplines are, therefore, necessary for an optimum and specific estimate of nutrition status. Both the potential and the shortcomings of a number of statistical techniques are demonstrated.