Browsing Mathematical Sciences - Masters by Research Theses by Issue Date
Now showing 1 - 7 of 7
Results Per Page
- ItemApplication of mixed-effects modelling and supervised classification techniques to public health data(University College Cork, 2019-09-28) Yang, Shuai; Fitzgerald, Tony; O'Sullivan, Kathleen (Catherine)This thesis consists of two parts. In PART A, we describe the application of mixed-effects modelling to 24 hour blood pressure. The blood pressure follows a 24-h circadian rhythm and the exaggerated morning surge in BP is an independent risk factor for cardiovascular diseases. In this project, the data analysed is from the Mitchelstown study. Morning SBP pattern between 4:00 am and 12:00 am was modelled using a piecewise linear mixed-effects model. Based on the likelihood function, the optimal breakpoint is at 7:30 am. Morning surge was characterised by the slope after the breakpoint. Model results revealed that the average slope between 7:30 am and 12:00 am is 2.47 mmHg/30 min (95\% CI: 2.35-2.59 mmHg/30 min). The Empirical Bayes estimates of subject-specific slopes were compared by age, gender, smoking, BMI, hypertension and diabetics. There were no significant differences in subject-specific morning surge between groups. Additionally, the relationship between chronic kidney disease (CKD) and the morning surge was explored using the multivariable logistic regression allowing for age, gender, smoking, BMI, hypertension and diabetics. Model results revealed that the association between the morning surge and CKD was not statistically significant. In PART B, supervised classification techniques are applied to SEYLE data. This project explores factors associated with drop-out in the SEYLE study. SEYLE study measured the mental health and wellbeing of adolescents with a baseline assessment and follow-up assessments at 3 and 12 months. Participant adherence is important when drawing inferences based on longitudinal data. However, drop-out in longitudinal studies are inevitable especially in adolescents. The primary objective of this project is to identify students with a high probability of drop-out in the SEYLE study using the Irish cohort. Multivariable logistic regression and decision trees (classification tree (CT), conditional inference tree, and evolutionary tree) were developed on a training data set. Factors considered included measures of sociodemographic, risk behaviours, lifestyle, general health, relationship and support, negative life events and psychiatric symptoms. Model performance was assessed on a test data set. Logistic regression analysis revealed that students aged 15/16, with chronic disease, normal anxiety level, high levels of hyperactivity, or lack of regular physical activity were significantly more likely to drop out of the SEYLE study. CT was regraded as the best tree and identified four subgroups based on age, anxiety and depression. Adolescents aged 15/16 without anxiety but with depression were classified as `drop-out' in this CT model. The choice between logistic regression and CT depends on the objective of the user. Logistic regression was the best at discriminating drop-out. However, CT is a simpler model and was marginally better at predicting drop-out.
- ItemActivity profiles of adults aged 50 - 70 years: functional data analysis(University College Cork, 2019-10) Weedle, Richard; O'Sullivan, Kathleen (Catherine); Fitzgerald, TonyPhysical activity has a major impact on health. Questionnaires are the most common method of physical activity assessment. While cost effective, these are subjective and can correlate poorly with actual activity levels. Accelerometers have gained popularity given their accuracy, objectivity and ability to capture large amounts of data. Simple summary measures such as the total or average activity over the day are often used. However, these fail to exploit the longitudinal nature of the data and do not capture the variation in activity levels throughout the day. This study intends to capitalise on this nature by implementing a functional data analysis approach. Activity data was collected from a cohort of 475 people in Mitchelstown in 2011. The individuals wore wrist worn accelerometers in a free living environment for a week. This data was collapsed into 1 minute epochs and each epoch was then aggregated over the week to get an estimate of daily circadian activity. The discrete wavelet transform was chosen as the smoothing technique to reveal the underlying functional nature of the data. This allows every individual in the cohort to be represented by a smooth activity profile. This study aimed to identify and characterise subgroups within a cohort based on these activity profiles. Functional principal component analysis was applied to these activity profiles in order to explore the dominant patterns within the data. Each individual’s profile was approximated by a weighted sum of profiles and these weights were then used to perform a cluster analysis. Five distinct subgroups were identified. These differed from each other in both the magnitude of the activity and the times at which the activity occured. A more simplified approach, based purely on the distance between profiles, was also implemented. Two distinct clustering methods identified the exact same 5 subgroups in the cohort. To ensure their robustness, these results were subject to a sensitivity analysis with respect to the epoch length, smoothing technique and number of functional components utilised in the clustering. Other studies have clustered accelerometer data in terms of absolute activity volume, as in high or low activity groups. However, they do not place too much value in using the granularity of the data to determine what time of day people are active. In addition to the high, moderate and low activity subgroups, our analysis revealed two subgroups which have a propensity to be active in either the morning or evening. It is suggested that these are indicative of an individual’s biological rhythm or chronotype. The Mitchelstown cohort was re-screened 5 years later in 2016, which presents an exciting opportunity to examine changes in these profiles over time.
- ItemA study of airline route data(University College Cork, 2020) Yu, Xiaochen; O'Sullivan, Janet; Wolsztynski, EricFor decades, with the advancement of airline information system construction, the aviation industry has successfully built a number of information systems. An enormous amount of data has been accumulated through the successful operation of these systems for the aviation sector. The effective use of these invaluable data assets has increasingly become a requirement for the relevant airline departments, and the focus of aviation industry. Revenue management is crucial for measuring the operational success of the airlines. However, the traditional forecasting methods cannot support processing of the underlying data that keeps changing over time, which have impaired the accuracy of the forecast, and thus, the credibility. The new airline passenger ticket revenue pricing methods proposed in this thesis have explored the possibility of solving the existing problems through advanced modeling techniques, and thus provide a view for better airline route planning and optimisation. First of all, the airline data is classified in a targeted manner. The factors of available seat kilometers, revenue passenger kilometers, load factors, total number of passengers, average fares, etc. collected from different time periods were used to establish a multiple linear regression model in the statistical software, R. Through empirical analysis it is found that the factors affecting the passenger ticket revenue, over different time periods of the same company, are different. Therefore, a multivariate linear regression model was established which was based on the data of different airlines in one specific time period. It was empirically found through this approach that different airlines had different factors affecting ticket revenue in the same time period. A multivariate linear regression model was established for analysing the data of the head office and the various branches at the same time period and through this it was found that the factors affecting the passenger ticket revenue of the head office and the branches at the same time period were different. The research results of this dissertation can provide scientific evidence that airlines should consider analysing real-time ticket data for ticket price and flight plan collected from the IT system to maximise the revenue.
- ItemA close knit complex network: statistical examination of board interlocks and their impact on financial results(University College Cork, 2021) Dave, Kirtivardhan Parantap; Roy, SupratikBoard of directors are at the pinnacle of the corporate world, in any company they form the highest decision making authority. A position on the board of a company is one of immense privilege and power. Thus, making any individual who occupies such a position privy to information about the company which most people would not have. Even the shareholders who are the owners of the company are not privy to this information. In this study I look at complex networks formed by board interlocks in the United Kingdom. When one director sits on the board of two companies, it creates a board interlock, where he acts as a link between the two companies. The main aim of the study is to look at the relationship between the features of these networks and their impact on the financial results of public limited companies in the United Kingdom. The first part of this study looks at these corporate networks and how they have evolved over the past three decades. Using varied network analysis features, the board interlock network was studied from two different perspectives. First, as a dual mode network or a people to company network. In this kind of network edges are formed between a board member and the company on whose board she sits. Second, as a single mode network where common board members between two companies form an edge between those companies. We find that the corporate network has become denser over the years, exhibiting small world properties. Companies that did not share board members with any other company have nearly halved over the course of the three decades. There have been quite a few studies that have looked at the impact of network features on financial results from various academic perspectives be it financial, sociological, management, and more recently network analysis. I have used four different centrality measures and proposed a ranking based model. Which unlike any other existing study also goes deeper by incorporating different community sizes into the regression model. This allowed me to check whether network features of better connected companies or companies that are a part of larger communities are stronger predictors of financial performance. This study shows an upward trend with the regression fit improving over time.
- ItemEmpirical analysis of relative impact of COVID-19 on sectoral stock returns in China(University College Cork, 2021-01-05) Chang, Keying; Huang, JianSince it broke out in China in December 2019, Covid-19 has brought an unprecedented impact on every aspect of social, economic, and cultural life globally. The direct impact includes a fragile system of health-care, cut off supply chain and trade, and sharp decline in production, consumption, and investment activities. The impact of Covid-19 on the stock market performance has attracted lots of research interests in the literature. In this study, we explore the relative effects of Covid-19 on the sectoral performance of the stock market in China, using daily time series stock market data spanning 1 st Jan 2020 to 30th Dec. 2020. First, we employed the principal component analysis to derive a proxy for each sector based on its composite shares. Second, we carried out t-test to estimate whether there existed a difference between average returns before and after Wuhan lockdown due to the outbreak of Covid-19. Finally, we employed the event study methodology to investigate the impact of Wuhan lockdown on the sectoral performance of stock market in China The event study found that Wuhan lockdown have significant effects on the stock returns in China. The pandemic has adversely impacted sectors such as Public Transport, Logistic Attractions and Tourism, Hotel and Catering. However, Manufacturing, Information Technology, Education and Health-care industries have been resilient to Wuhan lockdown.
- ItemRate-induced tipping in a moving habitat(University College Cork, 2021-03) MacCárthaigh, Ruaidhrí; Wieczorek, Sebastian; Hasan, CrisAs our planet's surface warms, its ecosystems are shifting to cooler and more suitable climates to survive . This issue raises some important research questions. How successfully can life on earth adapt to these changes? How fast can populations be pushed to migrate before they fail to adapt and collapse to extinction? This thesis addresses the problem of species adapting to shifting habitats, using the framework of ``critical transitions'' or ``tipping points''. Specifically, it explores the persistence of a single species in a fast-moving habitat, and how the nature of its population growth affects its adaptability. The spatiotemporal evolution of the species concerned is described by partial differential equations (PDEs) in the form of nonlinear one-dimensional reaction-diffusion equations and reaction-convection-diffusion equations. Two distinct growth models are constructed and the moving habitat is incorporated into each model with the addition of a continuous non-autonomous term. The first model is a quadratic monostable model of logistic growth limited by the carrying capacity of the habitat, similar to models used in previously published work on the topic of species in a moving habitat [2,3,4]. The second model is made cubic and bistable by incorporating an additional limitation to population growth at low population densities, known as the Allee effect. The spatial distribution of the populations are computed as standing wave solutions in the static habitat and travelling pulse solutions in the moving habitat, drifting at a constant habitat speed. A critical habitat speed is found for both the monostable and bistable models, above which no physical travelling pulse solution exists, meaning that the population within the habitat dies out. In the monostable Logistic Model, a transcritical bifurcation occurs between its travelling pulse solution and its zero (extinction) solution, which gives rate-induced tipping that can be reversed with a clear indication of an eventual local extinction as this critical speed is approached. In contrast, in the bistable Allee Model a saddle-node bifurcation of travelling pulses occurs, which produces an rate-induced tipping point, which typically cannot be reversed in nature, at which the population in the habitat abruptly collapses to extirpation, without any clear indication of the imminent collapse. This provides a stark ecological insight into how the increasing rate of climate change may give little to no advanced warning of extinction for some species that are observed to have an Allee effect and other complex features influencing their population dynamics.
- ItemStatistical applications in health and social research(University College Cork, 2022) Kenny, Alan; O'Sullivan, Kathleen (Catherine); Conway, Damian; Wolsztynski, EricThe MSc comprised four projects based on different statistical techniques. Each project is an individual chapter in the thesis and has been written independently. The first project is based on a student-wide survey conducted by University College Cork (UCC) to examine the student experience. As this study was unique to the Irish college system, factor analysis was employed to validate the use of purpose-built scales. Reliability analysis was conducted to determine if the scales were internally consistent, and the results of factor and reliability analysis and summary results were presented. The second project focuses on 15O Positron Emission Tomography (PET) Imaging Data and its ability to quantify underlying metabolic processes, namely cerebral blood flow in an Alzheimer’s disease patient. Kinetic analysis, using a 1-compartmental model, was employed and functional parameters were estimated. The technique was validated through comparisons with a non-parametric approach. This study confirmed that there was lower blood flow in brain regions that are affected by Alzheimer’s disease. The third project is an actuarial study, which explores the process known as graduation by standard table. Using information collected from a population, a set of crude rates are deduced. These rates need to be improved using the graduation by standard table method. Our graduated rates are then tested using five statistical tests to verify they are reliable. This study confirmed that use of the graduation method could improve crude rates to make them reliable. The fourth project focuses on structural empowerment in an Irish nursing context. Our study concentrates specifically on perceptions of formal power. Univariate analysis was conducted to identify associations between demographic characteristics and perception of formal power. Multivariate regression was used to identify which combination of demographic characteristics significantly affected a respondent’s perception of formal power. Respondents’ perceptions of formal power were in line with previous studies.