/library/oar/handle/123456789/6641 2025-11-01T17:02:54Z 2025-11-01T17:02:54Z /library/oar/handle/123456789/133771 2025-04-01T13:14:34Z 2024-01-01T00:00:00Z

Title: A multivariate Heston-Hawkes jump diffusion with application to high-frequency big tech stock prices Abstract: The tech industry has witnessed significant growth and disruption in recent years. Indeed, tech giants such as Apple, Meta and Microsoft play pivotal roles in the reshaping of traditional tech services. Investments in these tech giants are continuously evolving, where market dynamics are constantly influenced by real-time information. Understanding these dynamics is essential as these companies are rather significant in the market, due to their large market capitalisation. Furthermore, changes in the stock prices can signal shifts in technological trends and market perceptions of future developments. Taking this into consideration, this dissertation investigates the dynamic relationships between the stock prices of these big tech companies using a combination of the multivariate Hawkes process and the multi-asset Heston model. By utilising intraday data, the study explores the notion of jumps and volatility of an asset affecting not only the future values of the asset price itself, however also those for other assets. Theoretical foundations are laid out in chapters focusing on the d-dimensional Hawkes process, the d-asset Heston model, and a non-parametric jump-identification technique called the L-estimator. Furthermore, stock prices are analysed pairwise, by initially identifying the occurrences of jumps and fitting a bivariate Hawkes model on these jumps, followed by disentangling said jumps to obtain the continuous part of the data, in order to fit a two-asset Heston model on this data, thereby studying the stochastic volatility endured by the assets in question. The models fitted highlight that jumps are indeed mutually exciting, stochastic volatility is at its peak when the price is at its minimum, and that each of the asset price and volatility processes for AAPL, META and MSFT are moderately correlated, as anticipated. Description: B.Sc. (Hons)(Melit.)

2024-01-01T00:00:00Z /library/oar/handle/123456789/128127 2024-10-29T15:10:15Z 2024-01-01T00:00:00Z

Title: Analysing DASS through structural equation modeling Abstract: Structural Equation Modeling (SEM) is a statistical technique that investigates causal relationships between observed and latent variables. This dissertation provides the theory underlying the SEM framework, with specific attention directed to Muthén (1984)’s categorical variable methodology (CVM) that makes use of the weighted least squares (WLS) estimator to estimate the model parameters of a SEM fitted on observed categorical and ordinal data. SEM is applied to a dataset, obtained from Open-Source Psychometrics Project (2018) which consists of 42-ordinal-items related to the Depression, Anxiety, Stress Scale (DASS) where item value takes values from 1 to 4. The DASS is a clinical assessment tool composed of three self-report scales designed to measure the severity of symptoms common to depression, anxiety, and stress. One of the objectives was to use the DASS data to examine the mental health status in Europe focusing on four countries, namely, Germany, France, Great Britain and Poland. SEM was used to investigate any causal relationship between Depression, Anxiety, and Stress, and to examine the impact of gender and sexual orientation on these latent variables. Initially the study was conducted without considering the different countries present in the data. Subsequently, following testing for measurement invariance, a Multiple Group Analysis SEM was fitted, to determine whether the causal relationships between Depression, Anxiety, Stress, Gender and Sexual Orientation vary across the four countries. Interestingly, in France and Great Britain, LGBTQIA+ affiliation increased the likelihood of experiencing depression, less so in Germany, while in Poland, LGBTQIA+ individuals showed a lower likelihood of experiencing depression. Description: B.Sc. (Hons)(Melit.)

2024-01-01T00:00:00Z /library/oar/handle/123456789/128123 2025-03-14T07:11:18Z 2024-01-01T00:00:00Z

Title: A comparative analysis of hyperparameter effects on CNN architectures for facial emotion recognition Abstract: This dissertation investigates facial emotion recognition, an area of computer vision that involves identifying human emotions from facial expressions. It approaches facial emotion recognition as a classification task using labelled images from the FER2013 dataset, employing Convolutional Neural Networks for their capacity to process and extract hierarchical features from image data efficiently. This research utilises custom network architectures to conduct a comparative analysis of the impact of various hyperparameters—such as the number of convolutional layers, regularisation parameters, and learning rates—on model performance. Hyperparameters are systematically tuned to determine their effects on accuracy and overall performance. Notably, the best-performing model developed during this research surpassed human-level performance, established as being somewhere between 65% and 68% on the FER2013 dataset according to various studies. These findings provide a foundational understanding of hyperparameter optimisation for facial emotion recognition, demonstrating the impact of different configurations on model performance. Description: B.Sc. (Hons)(Melit.)

2024-01-01T00:00:00Z /library/oar/handle/123456789/126895 2024-09-24T10:38:01Z 2023-01-01T00:00:00Z

Title: Bayesian parameter estimation of the hurst index of fractional brownian motion Abstract: One of the many generalizations of Brownian Motion, Fractional Brownian Motion is very popular due to being able to account for a wide range of phenomena in various different fields ranging from finance when modelling stock data to hydrology in water turbidity analysis. Brownian Motion is unsuitable for modelling these due to the assumption of independence of increments, an assumption relaxed by Fractional Brownian Motion given it allows dependence of increments. This dependence is effected through the Hurst Index H, the parameter associated with the process. For values of H between 0 and 0.5, both excluded, negative autocorrelation between the increments is enforced while a positive one is obtained for values between 0.5 and 1, both excluded. As H approaches 0.5, the paths of the process will resemble those given by Brownian Motion and if equality holds, the process reduces to a Brownian Motion Process. In this thesis, the theory behind Fractional Brownian Motion as well as the Bayesian framework in the context of estimating H shall be discussed. Given the subjectivity involved in determining a prior, sensitivity analysis shall be performed as to analyze the effect of different priors on the posterior. Following this, the Hurst Index of real data shall be estimated. The data considered shall be genetic data relating to Covid-19 and cardiology data relating to heart rate variation. Description: B.Sc. (Hons)(Melit.)

2023-01-01T00:00:00Z