Browse
Recent Submissions
Item Penetration testing and vulnerability assessment on the Namibian inter-banking system: Namswitch(University of Namibia, 2023) Nghoshi, Sion S.Information Technology (IT) has become crucial to the operation of modern businesses. Financial service firms rely on IT to deliver services to their customers. In the Namibian context, the Bank of Namibia and the Payment Association of Namibia have discontinued the use of cheques as a payment method and provided Electronic Fund Transfers, cards, and electronic money as alternative payment methods. This means that the Namibian Payment System (NPS) now solely relies on IT in order to deliver services to its customers. The Namibian inter-banking system, known as Namswitch, is classified as an Information Service Provider to the NPS, enabling inter communication between Namibian financial institutions. Often, the target of cyberattacks are financial institutions and hackers with malicious intents are continually attempting to infiltrate their IT systems. As such, the financial services industry has unique information security requirements, and banks in particular conduct more stringent due diligence and due care in order to ensure the confidentiality, integrity and availability of their services. In order to address these security challenges, this study sought to explore ways to proactively strengthen and enhance the cybersecurity of the Namswitch system by evaluating the system’s security posture by proposing remedial actions, and further proposing a framework to automate and perform routine penetration tests in order to prevent future cyberattacks. The findings revealed the presence of vulnerabilities on the Namswitch system, some of which posed a high severity rating according to the CVSS risk rating. An example was the presence of default credentials on some internal systems and the use of low to medium strength ciphers on the external systems. A malicious user can leverage these vulnerabilities to perform attacks such as man-in-the-middle attacks. In an effort to strengthen the cybersecurity of the Namswitch system, the study provided a Namswitch Safe Financial Exchange (NAMSAFE) Protocol which is an algorithmic process aimed at remedying identified vulnerabilities and improving existing processes. It further outlines remedial strategies, risk mitigation steps, and compensating controls for vulnerabilities that could not be eliminated. Successfully implemented, NAMSAFE provides a prescriptive methodology for maintaining ongoing reliability and robustness to the Namibian banking systemItem A statistical analysis of endogeneity and instrumental variables of education on income in Namibia(University of Namibia, 2024) Nanyemba, Maria K.Education is a well-known driver of income and a causal factor for economic prosperity and social development in modern society. It provides one with capital such as qualification, knowledge and skillsets relevant to secure stable gainful employment, as well as increase one’s income, improve health conditions and general welfare. However, the measurement of the influence of education most often suffers from endogeneity suspicion due to the possible presence of a correlation between the education variable and the error term. This presents a critical issue when performing a cause-and-effect relationship analysis as neglecting endogeneity can lead to unreliable estimation results even when dealing with large samples because estimators of the model parameters will be inconsistent. To date, most studies that focuses on analysing the effect of education in Namibia have done so with other factors such as fertility and not income, while most of these studies used regression approaches such as the linear regression and logistic regression methods to perform their respective analyses. However, no study has simultaneously tried to account for the endogeneity of both education and income in Namibia. Thus, using the 2015/2016 Namibia Household Income and Expenditure Survey as well as accounting for the possible presence of endogeneity and for non-linear effects of continuous observed confounders, this study aimed at estimating the impact of education on income in Namibia, in addition to exploring different modelling strategies to capture the relationship between income and education, and other socio-economic and socio demographic factors. A total of 495,311 households containing 2,250,122 individuals aged 18 to 60 years who have been to school and have earned salary and/or wage as their source of income ii in the last 12 months of the survey period were considered in this study. Three different modelling strategies were carried out, namely: (i) Basic Ordinary Least Squares (OLS) regression without considering the clustering, stratification and weighting factors, (ii) Basic OLS regression considering the clustering, stratification and weighting factors, and (iii) Instrumental Variables & Two Stage Least Squares (IV-2SLS) regression, while the best fit modelling strategy to use for exploring the relationship between income and education, and other socio-economic and socio-demographic factors was determined using the Durbin and Wu-Hausman test of endogeneity. Results from this study showed that majority (82.1%) of the households in Namibia derived their source of income from salary and/or wage in 2015/2016. The IV-2SLS was identified as the best modelling strategy to use since the Durbin and Wu-Hausman test of endogeneity confirmed the presence of endogeneity (i.e, education must be endogenous), thus, making the OLS strategies unreliable to use. Furthermore, factors such as education, age, sex, marital status, type of work, employment period, total hours worked and transport mode to work had positive and significant influences on income, while factors such as region, dwelling unit type, main material for the wall, household size, and no contribution to investment had a negative and significant influences. However, the type of employer had a mixed effect on income. In addition, the IVs estimators revealed that the average rate of return of schooling was 13% which was higher compared to the OLS estimates of 5%. It is therefore recommended that the Namibian government and policy makers consider boosting more funding into education, especially in communities within the rural and underdeveloped regions still having traditional dwelling households whose structures are made of corrugated iron/zinc sticks, wooden poles, clay and/or cow dung. In addition, as part of their poverty alleviation strategies, the government should iii ensure the distribution of equitable resources, particularly the educational resources, to enhance the development of rural and underdeveloped regions so that they too can benefit from high-quality education opportunities in the countryItem Analysis of predator-prey models with infection in both species and human intervention(University of Namibia, 2024) Khoa, LutopuThe intricate interplay between infectious diseases and predator-prey dynamics holds pivotal significance. Within this context, our focus centres on the impact of human intervention through treatment and vaccination. This study delves into a predator-prey system, categorizing the populations as susceptible, infected with the disease, vaccinated when inoculated against infection, and under treatment post-infection. We formulate four comprehensive mathematical models that illustrate varying levels of human intervention: no intervention, intervention in the prey only, inter vention in the predator only, as well as intervention in both species simultaneously. Mathematical proofs of model positivity are provided. Following the derivation of equilibrium points, we analyse their stability by examining the signs of the eigenvalues from the Jacobian matrix and using the Routh-Hurwitz criteria. To verify our qualitative analysis findings, we conduct simulations using varied parameters in Matlab. We then draw conclusions regarding the impact that human intervention can have on a predator-prey system with infection. Simulation results indicated that without human intervention, predators faced extinction, whereas with treatment and vaccination in either the prey alone or in both predator and prey, the intervention demonstrated a positive effect, preventing the extinction of any species. Therefore, this study concludes that human intervention plays a crucial role in preventing species extinctionItem Reimagining the museum experience using augmented reality: A focus on a Namibian museum(University of Namibia, 2024) Chanakira, Tatenda AudreyThis study seeks to revolutionize the museum experience in Namibia through the implementation of AR Specifically, a focus was made on the National Museum of Namibia and developed an AR mobile application using ARKit and Reality Composer on the iOS platform. To gain a deeper understanding of traditional museum visits, an online survey was conducted which was targeted at various participants from different geographical areas, totaling 63 respondents. Background: The researcher's exploration of Namibian museums reveals the prevalence of two primary modes of visitor engagement: self-guided tours and guided tours, the latter often customized to suit the preferences and interests of international tourists. During personal visits, visitors typically rely on brief summary notes next to artifacts for supplementary information. Guided tours, while informative, often prioritize group interests, limiting individual exploration. Methodology: In this study, a mixed research design was employed to craft a software-based solution and gather insights from our target audience. By combining quantitative data from surveys with qualitative feedback from interviews and observations during AR application usage, we sought to gain a comprehensive understanding of user perspectives. Findings: This research unveiled that most museum goers had little to no prior exposure to AR applications during their museum visits, indicating an untapped potential for enhancing museum engagement through technology. The development and testing of the AR mobile application generated excitement among users, who expressed keen interest in additional features and interactive models. One challenge, however, was the application’s exclusivity to iOS devices, as it required iOS 13 or newer for full AR functionality. This limited the ability to collect feedback from a broader range of users. ii This study signifies a significant stride in reshaping museum experiences in Namibia. By leveraging Augmented Reality, the aim is to elevate visitor engagement, facilitate knowledge acquisition, nurture meaningful interactions, and forge emotional connections with museum exhibits. To extend the reach and impact of AR-enhanced experiences, exploring cross-platform compatibility to engage a wider audience is recommendedItem Analysis of malaria transmission dynamics in human and mosquito populations(University of Namibia, 2024) Shaanika, Hilma YvonneThis mini-thesis presents a deterministic mathematical model for the spread of malaria in human and mosquito populations. The human population is divided into four com partments while mosquito population is divided into three compartments. Suscepti ble humans can be infected when they are bitten by an infected mosquito, they then progress through the exposed, infected, recovered before going back to the susceptible class. Susceptible mosquitoes can be exposed to the disease and once they are exposed, they can be infected, and remain infected until they die. Basic reproduction number, R0 was established and used to determine whether the disease dies out or persists in the population. It was shown that the disease-free equilibrium point is locally asymptoti cally stable when R0 < 1 and unstable when R0 > 1. Quantitative analysis of the model was carried out to confirm the findings from qualitative analysis. Result obtained indi cate that the findings of quantitative analysis correspond to the findings of qualitative analysis. It was proven qualitatively that R0 < 1, which corresponds to the results of the sensitivity analysis, that was carried out quantitatively. It was recommended that future work can be done to investigate the stability of the endemic equilibrium pointItem Modeling risk factors for psychotic conditions and depression among adults in Namibia(University of Namibia, 2019) Panduleni, Haufiku AdolfMental disorder is a growing global health concern. Mental disorder is categorized into anxiety, depression, schizophrenia, alcohol and drug dependency. Depression is ranked by the WHO as one of the largest contributors to global disability and suicide. There are only two affordable mental health services in Namibia, this leads to underdiagnosed and untreated cases of people suffering from depression and psychotic disorders. The overall objective of this study wasto identify socio-demographic factors associated with psychotic disorders and depression among adults in Namibia. The specific objectives of this study was to: estimate the prevalence of psychotic disorders and depression among adults in Namibia, construct an index of depression and identify factors influencing psychotic disorders and depression among adults in Namibia. This study adopted a quantitative cross-section study design, based on secondary data from the NDHS 2013. The study used a sample of 14234 participants, 9906 women aged 15 – 49 and 4328 men aged 15 – 64. Binary logistic regression was used to establish factors contributing to psychotic disorders. Factor analysis (Principal component analysis) was used to construct an index for depression and grouped the single factor into terciles. Then multinomial logistic regression was used to establish factors influencing depression. The results indicated that psychotic disorder had a prevalence of 13 percents while depression had 30 percents prevalence, in 2013. The psychotic disorder was found to be significantly associated with region, educational level, marital status and sex using binary logistic regression. Multinomial logistic regression results indicated that both mild and severe depression conditions were found to be significantly related to the region, smoking, marital status,sex, and employment status. Moreover, smoking cigarettes and place of residence were significantly associated with mild depression. There is a need for more mental disorder services to be promoted and rolled out to all health facilities in Namibia. It is also emphasized that there should be mental health education campaigns in primary schools and communities in the country. Marital relationships (marriages) should be promoted in society. Smoking should continue to be discouraged by increasing the price of cigarettesItem Vulnerability assessment of information systems based on end-user actions: A case of University of Namibia(University of Namibia, 2022) Kautwima, Paulus S.Nowadays, data protection is of paramount importance to every institution of higher learning. Unfortunately, most security breaches happen as a result of an end-user error(s). End-user errors are unintentional actions of system end-users. This includes; weak passwords, clicking links from unverified senders, and downloading attachments from unknown senders. Most of the efforts aimed to address cybersecurity issues are either software-centered or hardware oriented. As a result, user mistakes are overlooked since they are considered infinite, unpredictable, and remain part of human existence. Ignoring end-user error is a gigantic mistake and could result in a higher number of cyber-attack incidents. Therefore, institutions of higher learning where security is a top priority need innovative strategies to deal with end user errors. Given this, this study aimed to assess different types of end-user errors that could affect the security triad of information systems. The study adopted the mixed-method research approach to collect data from the University of Namibia (UNAM) staff members, who frequently use information systems known as ITS. The quantitative dimension of the study utilized a closed-ended questionnaire to collect data from 310 UNAM staff members, who were randomly selected from the total population. Furthermore, an experimental design was also used to collect data from the staff members. The qualitative dimension utilized an exploratory research design where participants were selected through a purposeful sampling strategy. A semi-structured interview instrument was also applied to collect data from 10 staff UNAM Computer Centre staff members. The findings of the study revealed that end-user error is one of the major threats to information security. End-user errors present several security vulnerabilities and risks to information systems that could subsequently get data exploited by attackers. In addition, the study also established that confidentiality, integrity, and availability of information systems in an institution are also affected by end-user errors. Furthermore, the unprecedented growth of internet interconnectivity has led to an enormous increase in cyber attacks. Personal security consciousness and security awareness training are some of the most successful measures to mitigate end-user errors. Based on the findings of the study, it is recommended that institutions enforce information security policy and provide security awareness training to staff members to avoid data breachesItem Determinants of women’s participation in Namibia ’s labour force: A multinomial analysis of the 2018 Namibia labour force survey(University of Namibia, 2023) Amulungu, Fenny MukwiilongoWomen’s participation in the labour force is regarded as a sign of decreasing discrimination and a slight increase in women empowerment, among other benefits. Various studies have been conducted worldwide on identifying factors associated with women’s participation in the labour force but not many studies have been conducted on this topic in recent years in Namibia besides Mufune (2013). This study performed an empirical investigation to identify determinants of women’s participation in Namibia’s labour force through the adoption of a quantitative cross sectional research study using the 2018 NLFS and a multinomial logistic regression technique. Results revealed that area location, region, age group, marital status, literacy status and education level were significant determinants of employed women’s participation in the labour force in Namibia, while area location, age group, and literacy status were significant determinants of unemployed women’s participation. Compared to the odds of economic inactiveness, women from Hardap (OR=0.697, p<0.001, 95%CI: 0.667-0.729) and Kavango East (OR=0.921, p<0.001, 95%CI: 0.885-0.959) regions, relative to those from Zambezi region, had low odds of employment. Women who were less than 20 years old (OR=0.106, p<0.001, 95%CI: 0.102-0.109), relative to those aged 60 years and above, had low odds of employment. Married women (OR=0.711, p<0.001, 95%CI: 0.666-0.759), relative to separated women, had low odds of employment. Moreover, women with no education (OR=0.200, p<0.001, 95%CI: 0.186-0.216), those with primary education (OR=0.254, p<0.001, 95%CI: 0.236-0.273), junior secondary education (OR=0.288, p<0.001, 95%CI: 0.268-0.309), senior secondary education (OR=0.338, p<0.001, 95%CI: 0.315-0.363) and technical/vocational certificate/diploma (OR=0.262, p<0.001, 95%CI: 0.240-0.286), relative to those with postgraduate ii certificate/diploma/degree, had low odds of employment. However, literate women (OR=1.636, p<0.001, 95%CI: 1.594-1.679), relative to illiterate women, had high odds of employment compared to the odds of economic inactiveness. Furthermore, women who were from the Erongo (OR=1.132, p<0.001, 95%CI: 1.085-1.182), Kunene (OR=1.572, p<0.001, 95%CI: 1.499-1.648), Omaheke (OR=1.456, p<0.001, 95%CI: 1.384-1.531), Oshana (OR=1.098, p<0.001, 95%CI: 1.054-1.143), Oshikoto (OR=1.136, p<0.001, 95%CI: 1.091-1.182) and Otjozondjupa (OR=1.631, p<0.001, 95%CI: 1.562-1.703) regions, relative to those from Zambezi region, had high odds of unemployment as compared to the odds of economic inactiveness. Likewise women aged 20-29 years (OR=47.014, p<0.001, 95%CI: 44.570-49.593), 30-39 years (OR=66.831, p<0.001, 95%CI: 63.295-70.564), 40-49 years (OR=38.702, p<0.001, 95%CI: 36.648-40.871) and 50-59 years (OR=11.415, p<0.001, 95%CI: 10.799-12.065), relative to those aged 60 years and above, had high odds of unemployment. Women in consensual union (OR=2.124, p<0.001, 95%CI: 1.931-2.337), relative to those who were separated, had high odds of unemployment, while those residing in urban areas (OR=0.869, p<0.001, 95%CI: 0.853-0.885), relative to those from rural areas, had low odds of unemployment. It is therefore recommended that the Namibian government as well as policy makers and implementers reinforce policies and legislative frameworks in place to (further) empower women and enhance their participation in the labour force and in return, reap the benefits of economic and social developmentItem A bayesian hierarchical modelling of small area variation in youth unemployment in Namibia(University of Namibia, 2023) Shitenga, Linda VuteYouth unemployment has been one of Namibia's socio-economic problems, which has the potential to have significant and serious social repercussions on economic growth and development and could cause social exclusion and unrest in the affected country. Youth unemployment rates estimates are only available at the national (46.1 percent in 2018) and regional levels in Namibia; however, the Namibia Labour Force survey (NLFS) does not provide such statistics at small area e.g., at constituency level. The census data could be used to provide estimates of youth unemployment at constituency level; however, the data is only obtained every 10 years which in most cases the time frame is too long given the developmental changes that may take place during the 10- year periods. In view of these challenges, it is paramount to estimate the variation in unemployment rates at constituency level for possible targeted interventions within regions. In comparison to conventional small area estimation (SAE) models, the hierarchical Bayesian approach to SAE problems has several benefits, one of which is the ability to properly account for the kind of surveyed variable. For this reason, the main objective of this study was to estimate the risk of youth unemployment at constituency level using the 2018 NLFS data. The likelihood was estimated using a hierarchical Bayesian model and results from the study showed that the chance of youths being unemployed was very high among male youths than female youths in urban areas with OR=1.35 (1.10, 1.66) and OR=0.79 (0.65, 0.96) respectively. Several models were fitted, and the best model was used to estimate the probability of being unemployed amongst the male and female youths (with the DIC values of 4900.90 for the males and 5719.48 for females). The best model considered the fixed effects together with the unstructured spatial effects at constituency and regional levels. Even though employment is the result of aggregate demographic and socio-economic factors, the study recommends that employment opportunities specifically targeting youths should be created by either government or private sectors in constituencies especially those in the rural constituencies. Furthermore, special attention should be paid to integrating the youth into the labour market by improving their educational levelsItem A comparison of threaded versus non-threaded computing on the sieve of Eratosthenes Algorithm(University of Namibia, 2023) Uushona, SuamaPerformance enhancement in computers is a constant challenge for computer engineers, emanating from the constantly changing needs of consumers. Around the 1970s and 80s, computer engineers started designing computer architectures with multiple processors onboard, in order to address the performance challenge. These newer architectures provided a platform that enabled multiprocessing at affordable retail prices. This in turn created an opportunity for software developers to enhance program performance by creating software that could leverage on the underlying architecture through parallelisation to provide a performance advantage over sequential programs. This was not always achieved, however, since parallel programs present more opportunities to generate overhead, which can limit, or even deteriorate a program’s overall performance. In this study, an experimental analysis of two Sieve of Eratosthenes programs, one sequential and one parallel, was conducted in order to verify that parallelisation provided a computational advantage in the program under investigation, to establish whether a variation in the number of available processors had an effect on the overhead incurred, and lastly to investigate the mathematical nature of the overhead incurred. The study concluded that the parallel program provided a computational advantage over the sequential program for all threads computing prime numbers in the ranges greater than 10,000. The results of the study also determined that there existed a positive statistically significant relationship between the number of threads employed and the overhead incurred. In addition to this, the study also determined that overhead was mathematically quantifiable, but not in relation to the number of threads employed. Instead, it was learned that overhead is a product of the parallel program’s execution time and the fraction of efficiency lost. All in all, the study certainly highlighted one benefit of parallelisation, namely performance enhancement.Item Exploring the awareness of security threats associated with short-message service (sms) and protective measures against sms security threats amongst students at the University of Namibia (UNAM)(University of Namibia, 2023) Kalipi, AbrahamThis study focussed on exploring the awareness of security threats associated with short message Service (SMS) and protective measures against SMS security threats amongst students at the University of Namibia (UNAM). Preliminary interviews with six (6) students with different demographics, informed the design of a survey of students’ encounters with, perception of and responses to SMS security threats. Two hundred and eighty-seven (287) students registered at UNAM’s twelve (12) campuses countrywide in the 2019 academic year responded to a questionnaire that comprised of forty questions/sub-questions about participants’ demographics, experiences and perspectives on security threats associated with the use of SMS. Data were analysed using descriptive and correlation statistics and structural equation models. Participants had some awareness of SMS security threats and used basic security measures such as passcodes, patterns, and biometric access. However, most did not have extensive knowledge of the types of fraudulent activities such as phishing, DoS attacks, relay attacks, and spamming, nor were they aware of protective measures to counter these threats. Only 4% of participants had installed antivirus software on their phones. Junior students comprised 77.7% of participants, which might explain the overall low awareness. Participants indicated that some organisations such as Namibian banks promoted their awareness of security threats, and they thus suggested that the university should run awareness programmes and campaigns. Based on insights about participant demographics and relationships to the university, recommendations were made about how a university might increase students’ awareness of some protective measures. These include integrating a compulsory course on cyber security awareness into the university’s curriculumItem Multilevel analysis of academic performance in grade 12 STEM subjects, in secondary schools in Namibia(University of Namibia, 2022) Komotolo, EsterAccording to Francesco and Nicole (2015), the fourth goal of the United Nations Sustainable Development Goals (SDGs) aims “to ensure inclusive and quality education and promote lifelong learning for all learners to have access to education” by 2030 to facilitate literacy and numeracy equal opportunities. High-quality STEM education is a global prerequisite for individual development and participation in a technology-driven world, no wonder the need to focus on education programs on science and technology. Skills such as critical thinking, problem solving and the ability to innovate are increasingly important for openly embracing change and responsibly shaping the future. The aim of the study was to examine the relationships between academic achievement and demographic variables and socio-economic variables and to what extent students’ factors; teacher’s factors and school factors affect the students’ academic performance in STEM subjects in secondary schools in Namibia. Using a hierarchically built data of 24407 students nested in 190 schools and a series of two-level multilevel model were explored to determine predictors of academic performance. The AIC established the balance between accuracy of model and its complexity. The model that had the lowest AIC was considered the best. The Intra Class Correlation (ICC =0.169) indicated remarkable clustering of the number of STEM subjects passed within centres. The findings indicated that, the school type were not significant to the study. Teachers with and without formal education were found with a negative significant toward STEM subjects The study further established that, number of classrooms, number of media or resource centres, students’ age and students’ ii sex, were found to be significantly related to students’ STEM subjects passed. Based on findings, the study recommends that, there is a need for a comprehensive concept that addressesseveral dimensions of proper coordination and management school environments to ensure that students are equipped with the necessary resources in terms of large numbers of students in classroom, learning facilities and ensure qualified teachers are employed.Item A multilevel logistic regression analysis on suicide ideation and associated factors among school going adolescents in Namibia(University of Namibia, 2022) Kapenda, Sesilia Rauha N.Suicide accounts for many deaths worldwide making it the second leading cause of death among 15-29 year olds globally. Suicide ideation leads to mental health disorder and has been shown to “spread” suicidal thoughts to others. Consequently, adolescents who reported suicidal ideation at an early age are more likely to later attempt suicide. The objective of the study was to estimate prevalence and risk factors for suicidal ideation among school-going adolescents in Namibia. The study analyzed secondary data from the Namibia Global School-based Health Survey (GSHS) of 2013 (N= 4 531). A multilevel logistic regression, which allows for the incorporation of group level effects (school and grade) and estimation of interactions, was used to establish risk factors associated with suicidal ideation. 32% of the school-going adolescents indicated they had considered suicide at least once in the previous 12 months. Age (p = 0.025), sex (p = 0.025), dietary behaviour (p = 0.035), violence (p-value < 0.001), mental health (p = 0.002), drug use (p = 0.008) and tobacco use (p-value < 0.001) were significantly associated with suicide ideation. Results suggest significant school level effects with a VPC of 0.0454 and grade level effects witha VPC of 0.0231. Due to a small number of grades and schools, the maximum likelihood estimation may have produced unstable estimates and therefore, future researchers may use a Bayesian approach. Future researchers can also make use of multilevel multinomial logistic regression modelling instead of the multilevel binary logistic regression modelling that was used in this study.Item Measuring youth poverty in Namibia: An application of a multidimensional, multilevel modelling approach(2022) Shifotoka, Selma N.M.Officially, poverty in Namibia is measured by means of monetary thresholds, using the World Bank’s Cost of Basic Needs (CBN) approach. Poverty cannot, however, be solely defined by the lack of monetary resources. Rather, it is a combination of a range of non monetary factors which act as constraints on individuals’ abilities to reach their capabilities. Adopting the Oxford Poverty and Human Development Initiative (OPHI)’s Multidimensional Poverty Index (MPI), an index that captures acute deprivations that a person faces simultaneously, this study assessed multidimensional poverty rates amongst the youth (15-34 years) in Namibia. It examined the incidence and intensity levels, as well as the determinants impelling the level of youth poverty at individual, household and regional levels. The results from this study indicate that the prevalence of youth multidimensional poverty in Namibia stands at 31.4 percent. Across demographic groups, the results show that multidimensional poverty was high amongst females (32.2 percent), the younger youth aged 15-19 years (43.1 percent), those who resided in rural areas (42.8 percent), as well as those who lived in households that were headed by females (33.7 percent). The dis-aggregation of multidimensional poverty measures by regions indicates that the three regions with the highest rates were Kunene (56.1 percent), Kavango West (53.3 percent) and Kavango East (50.4 percent). The intensity levels further showed that the regions with the most deprived youth also had the most severe poverty. The study found significant determinants of the prevalence and intensity of youth multidimensional poverty not only at the individual level, but also at household and regional levels. In addressing youth multidimensional poverty, the study recommends an integrated approach that takes into account the hierarchical socio-economic effects on the livelihood of the youth, strengthening female’s integration into the labor market with equal access to social protection and equalizing rural and urban opportunities in the labor, health and education sectors.Item Evaluation of machine learning classification models for detecting electronic fund transfers scam SMSes(University of Namibia, 2020) Enkono, Fillemon SimanekaThe last decade saw the emergence of mobile banking and a pervasive transcendence of spams from email to SMS communications. M-banking offers the users an ability to execute EFT transactions using mobile devices and allow them to receive SMS notifications acknowledging their transactions. While this provide convenience to m-banking users, in the wake of SMS spams it also presented vulnerabilities that could be exploited to scam money and goods from them. To execute these scams, spammers send forged EFT ( e.g. e-wallet) deposit notification SMSes to unsuspecting users, then contact and request them to do EFT payments as refunds for the supposed en-oneous deposits acknowledged by the bogus notifications. Similarly, during goods exchange, scammers use forged deposit notification SMSes to trick sellers to believe that they have paid for the goods. In Namibia, the high affordability of SIM cards and the readily available access tom-banking accounts such as e-wallet by anyone with a valid SIM number provides a favourab le operating environment for the EFT SMS scammers. Inferences from literatures on novel spam filtering techniques suggested that implementing machine learning classification could help address the EFT SMS scams problem, partly motivating this study to evaluate such application. Prevalent reporting of EFT SMS scams in local media (which mostly involves the country"s largest bank by market share, FNB) and the observed lack of dedicated IT solutions to address such problem were other factors that inspired this work. The study collected a dataset of ham and EFT scam SMSes, from which machine learning features for classifying SMSes were extracted. This was followed by a pre-evaluation to determine the features that allow ham and EFT scam SMSes to be classified optimally. SMSes comprising the collected dataset were then represented using the optimal features and used to train and evaluate Suppo1i Vector Machine, Naive Bayes and Random Forest classifiers. The evaluation results revealed that the SVM classifier was the most effective with respect to detecting EFT scam SMSes, achieving a FNR=0.00, CA=0.992, Recall= l .0 and F l-measure=0.995. The RF classifier followed with FNR=0.0 11 , CA=0.983, Recall=0.989 and F l -measure=0.989; while the NB classifier came last with FNR=0.027, CA=0.975, Recall=0.973 and F l -measure=0.983. The envisaged future work will look to use the methods, findings and conclusions drawn in this study to guide development of mobile application(s) that implement machine learning classification to detect EFT scam SMSes.Item Neighbouringhood distinguishing colourings of graphs(University of Namibia, 2020) Nafuka, Helena NIn this mini thesis, we study neighbourhood distinguishing colouring (NDC) of graphs, which are proper colourings of the vertices with the added condition that for every pair u, v of distinct vertices there is some colour c such that the number of vertices of colour c adjacent to u is different to the number of vertices of colour c adjacent to v. The neighbourhood distinguishing colouring number XNDc( G) is defined as the minimum cardinality of a neighbourhood distinguishing colouring of a graph G. The study begins with the discussion of some terminologies and definitions used later on in our study. Moreover, we consider the colour classes corresponding to an NDC and the neighbourhood distinguishing colouring number of certain familiar classes of graphs such as paths, cycles and trees. In addition, we classify graphs with neighbourhood distinguishing colouring number XNDc( G) equal to two up to isomorphism. The chromatic number Xa of graphs G with XNDC equal to two is also two. Finally, we characterize graphs whose XNDC coincides with the order of the graph. These graphs possess a unique•XNDc- partition and they are either complete graphs or union of vertex disjoint edges. A XNDc-partition of a graph G is a partition of G with XNDC elements. The aim of this study is to give a considerable discussion of the neighbourhood distinguishing colouring and also to light the way for further research in the field of colourings.Item A comparative evaluation of unsupervised anomaly detection techniques in smart water metering networks(University of Namibia, 2020) Kanyama, Maria NelagoAdvances in electronics and wireless communication technologies have enabled the development of smart water metering networks (SWMNs). These networks enable water utilities to measure the water consumption of the connected households better than the traditional metering. Every household's traditional water meter is replaced by a smart meter. Smart meter communicates the household's water consumption to the utility. Data collected by the smart meters can easily become unreliable due to different data attacks such as data tampering, integrity attacks or flooding. Data integrity has become a major concern in SWMNs hence this work is motivated by the demand for data integrity in this networks. Thus, different security algorithms and techniques needs to be put in place in order to detect which smart meter is compromised. Anomaly detection techniques are some of the security techniques that can be applied in SWMNs in order to protect all SWMN stakeholders. Although a lot of privacy and security solutions for attacks are proposed in smart grids, there is a dearth in security solutions for SWMNs in literature. This work is a comparative evaluation of three unsupervised anomaly detection techniques in Smart Water Metering Networks (SMWNs). These techniques are k-Nearest Neighbor (kNN), cluster-based local outlier factor (CBLOF), and the histogram-based outlier score (HBOS). The comparative study aims at providing a better unsupervised anomaly detection technique that can be adopted in SWMNs. This work aimed to find a better anomaly detection technique that can be used in a SWMNs, to track down the smart meters that are producing anomalies, the time the anomaly occurred, and lastly, to plot the consumer's profiles. Simulations are conducted in OMNeT ++ v5 .0 platform running on the Ubuntu 14.04OS. INET-3.4.0 framework is used on top of OMNeT ++ v5 .0 in order to simulate and model a realistic network. Data collected from the Tsumeb East area is used to configure the normal meter reading. Several scenarios depicting deviations from the normal conditions are simulated. The performance of each technique is compared with one another. False positive rates (FPR), detection rate (DR) and accuracy rate are used as performance metrics. One-way Anova and Turkey HSD are used for statistical analysis. Simulations runs for 60s; the first 20s aim to build the intelligence for the algorithm and the remaining 40s as a test phase. Simulation results show the kNN achieves almost zero FPR throughout while giving almost 5% for DR in all simulation runs. On other hand, CBLOF achieves the highest DR between 95% - 100%, while giving the worst performance in terms of FPR. This shows that CBLOF's anomaly score is too stringent as it penalizes even normal data points. The HBOS exhibits intermediate performance in either metrics. HBOS gives out the intermediate results. Statistical results for accuracy rate show that although HBOS gives intermediate results, it tends to run very close to CBLOF. kNN gives a high accuracy rate while CBLOF gives a lower accuracy rate since it struggles to identify normal data. This works concludes that kNN is a better anomaly detection technique to be used in SWMNs because it has a better DR, low FPR and a high accuracy rate as long as the magnitudes for the anomalous readings are not too close to what is being considered normal. Future work will explore a hybrid anomaly detection technique. Future work will also look at how to distinguish anomalous data from normal changes in customer water usage habits.Item Mapping the human immunodeficiency virus/acquired immunodeficiency syndrome epidemic in Namibia using Bayesian spatial modelling(University of Namibia, 2020) Shikongo, JobIn order to develop an effective prevention response to the Human Immunodeficiency Virus/Acquired Immunodeficiency Syndrome (HIV/AIDS) epidemic in a particular area, a deeper understanding of the dynamics of the epidemic is required. Disease mapping is usually used to determine the spatial distribution of a disease prevalence and its risk factors in a particular area in order to device appropriate interventions. Although maps of HIV transmission are generally needed for planning, resources allocation, and monitoring and evaluation, such maps were currently not available in Namibia. This study developed a spatial model of HIV/AIDS epidemic in Namibia based on Bayesian methods (spatially unstructured and structured random effects) using the 2013 Namibia Demographic and Health Survey data. Furthermore, the study identified socio-economic demographic characteristics and sexual behavior that were associated with HIV/AIDS prevalence in Namibia. Specifically, spatial regression models were fitted using BayesX 3.0.2 to adjust for spatial random effects and non-random effects, and the Moran’s I statistic was calculated to test for the significance of autocorrelation between neighboring regions to show if they tend to cluster. The Moran's I statistic (0.120) was significant (p-value = 0.003) with a variance of 0.002 which stipulated that values that determine the strength of spatial dependence in neighboring regions tend to cluster. After adjusting for spatial random effects and non-random effects, results shows significant structured spatial effects with posterior mean ranging between (-0.423, 0.759) at regional level and (-0.687, 0.995) at constituency level. The socio-economic, demographic and cultural factors like non - condom use, wealth index (poor, middle, richer), marital status (living with partner) and gender (male) were significant in explaining the HIV prevalence in Namibia. Spatial clustering was observed in Khomas, Erongo and towards the regions in the northern parts of Namibia, namely Oshana and Ohangwena. The study recommends that the modelling of relative risk (as a function of spatial structure and spatial instructed random effects) in Namibia using Bayesian multi-scale models should be based on census data in order to identify definite spatial structures which would be exceedingly critical for both illustrative as well as policy implementation purposes.Item Mathematical modeling of foot and mouth disease (FMD) in Cattle and Buffaloes using vaccination and culling: A Namibia perspective(University of Namibia, 2022) Shikumwifa, EvalthineFoot and mouth disease (FMD) is an acute, infectious viral disease for animals, and it is one of the most rapidly spreading diseases worldwide. Countries worldwide are putting efforts to curb the infection as it has devastating effects on agriculture and wildlife economies. Mathematical models have been used to analyse the transmission and control of FMD to enable better decision making for animal health policy makers. In this study, we developed and analysed a basic mathematical model of the dynamics of FMD with and without vaccination and culling as control measures replicating the FMD infection in the interface setting of Namibia communal areas and National parks. Furthermore, we fit the model with control measures to the yearly cumulative FMD cases in Namibia and discuss the results in order to identify the impact of these controls on the Namibian scenario. Vaccination and culling are the control methods mostly used in the Namibian setting for the control of FMD. Mathematical theories for systems of ordinary differential equations were used to establish the existence and uniqueness of model solutions, as well as the stability of equilibrium points and to ensure that the mathematical solutions were biologically reasonable. Important threshold parameters such as the reproduction numbers were established, which are critical indicators of disease spread. Results from the study showed that the models have two equilibrium points namely; the disease free equilibrium (DFE) and the endemic equilibrium points. The DFE was shown to be locally asymptotically stable when R0 < 1 and unstable when R0 > 1. The endemic equilibrium point was shown to exist when R0 > 1 and globally stable when R0 > 1. Numerical simulation results, using Namibia data relevant to the transmission dynamics of FMD, were presented to illustrate some of the main theoretical results and model projections. The results from this study suggest that increasing vaccination rate and efficacy has a positive impact on reducing the spread of FMD. Moreover, we have also observed that vaccination is not enough to protect both the cattle and buffaloes, therefore, more efforts from policy makers should be devoted to putting extra measures in place for buffaloes and cattle not to interact more often. Better results were observed when both vaccination and culling were implemented, hence it is advisable to Namibia to practice culling as one of the control measures.Item An ecological adjusted random effect model for violent and property crime in Windhoek (2011-2016)(University of Namibia, 2020) Amunyela, JonasCount data that are zero inflated are often analysed using Zero-Inflated Negative Binomial Generalized Linear Mixed Model (ZTNB-GLMM) when observations are correlated in ways that require random effects. The present study investigates ecological factors influencing Violent and Property crime in Windhoek by using data obtained from the Windhoek police over the period of six consecutive years (20 I I to 20 16). The ecological concepts were measured at several different levels of aggregation. Limited studies in Windhoek have considered analysing crime data on a newly established Generalized Linear Mixed Model via Template Model Builder (TMB) R-package. The researcher considered the number of reported Property and Violent crime for the study period as a quantitative design. Crime was counted with respect to Month, Season, Year, Location and Density. Through an exploratory study, it was found that both Property and Violent crime data contained more zeros than would be expected. Furthermore, in specifying the probability distribution using confidence interval, the researcher found out that the Negative Binomial distribution was appropriate for the two types of crime. Besides that, the lognormal distribution also appears to be an appropriate distribution for modeling Violent crime. However, when comparing models fitted in the context of these two distributions it was found that the Relative Risk (RR) were highly significant for models fitted via Negative Binomial distribution. By adopting a ZINB-GLMM, the study attempts to address the potential covariates for both Property and Violent crime. The study shows that most of the variation in the study of Property and Violent crime was due to locations. On average more Property crime (68%) was committed in Windhoek as compared to Violent crime (32%). Crime was high during Spring and Winter time during the study period. The study further discovered that areas with high population densities have a high crime intensity.
- «
- 1 (current)
- 2
- 3
- »