/library/oar/handle/123456789/107554 2025-12-28T04:09:07Z /library/oar/handle/123456789/141986 Title: Bayesian approaches for ligand-based virtual screening applications Abstract: The objective of computer aided drug design is to discover new drugs by carrying out algorithmic modelling of chemical interactions of bioactive molecules. Drug discovery is known to be a notoriously lengthy and costly procedure, therefore this sparks a great motivation for further research in the field to be carried out in order to simultaneously reduce the time elapsed during drug discovery and also produce effective products. Virtual screening is an umbrella term for a variety of ligand-based and structure-based tools which are used to search databases of chemical structures. Of particular interest to this study is ligand-based virtual screening and this uses known and active compounds to a specific target to screen molecules of unknown activity. We explore the way in which statistical approaches, specifically Bayesian statistics have been adopted for Ligand-based Virtual Screening. We implement two main similarity models, the Bayesian Inference Network and the Bayesian Belief Network, as well as explore model tuning avenues as an attempt to improve upon our results. The first crucial research question we seek to answer is if such statistical approaches provide better screening results when compared to conventional similarity scoring techniques, specifically the Tanimoto similarity metric. Indeed, through our research we show that the Bayesian similarity models developed through this dissertation do in fact improve screening results. Significant improvements in the ROC AUC is recorded when the Bayesian Inference Network and the Bayesian Belief Network are employed instead of the Tanimoto similarity metric, with maximum improvements of 15.52% and 15.19% respectively. Secondly, we aim to determine whether screening effectiveness is improved when multiple actives to a known target are used to rank a compound database as opposed to a single active. Through this research we suggest that for such Bayesian similarity models for Ligand-based Virtual Screening, a single active provides better results. Description: M.Sc.(Melit.) 2022-01-01T00:00:00Z /library/oar/handle/123456789/141774 Title: Managing pain through federated affective computing Abstract: For a long time, the primary approach to control pain in patients involved using specially designed drugs. While these drugs have often proved to be sufficient to reduce pain perception in patients of all ages, they do not come without any potential side effects. The prolonged use of such drugs can have adverse effects on a patient's health, namely through the development of tolerance and resistance to medication. This research explores a shift away from such practices. It proposes using technology as a safe adjunct tool that, together with a reduced need for specialized medication, helps patients cope with pain. This study explores the adoption of Affective Computing as a crucial element in administering effective Distraction Therapy to shift patients’ attention away from pain and onto more pleasant thoughts. Described as the study of applying human-like capabilities to machines, affect-enabled software can elicit, understand, and, more importantly, react to human emotion. While interest in the field of Affective Computing has grown considerably over the years, the implementation of such software is often faced with obstacles that threaten the successful creation of this kind of software. Affect-enabled software, as with any other kind of software requiring a degree of Machine learning, can be considered highly data-hungry. In closed-loop Affective Computing implementations, significantly large bodies of data train accurate models upon which they make informed decisions. Among the list of issues that have for long hampered the development of intelligent but data-hungry software is the lack of sufficiently large data collections. In addition, given that one is dealing with affect, training one static model to be used by multiple individuals might not return the ideal results because, as humans, we react in different ways when faced with different scenarios. Thisresearch presents the Federated Affective Computing concept. Federated Affective Computing brings together Affective Computing and Federated Learning to overcome the lack of data hampering the development of affect-enabled platforms. The result is a framework that, while overcoming such issues, makes possible the creation of affect- enabled platforms that can autonomously continuously retrain their models to remain relevant. By introducing Federated Learning using Evolutionary Aggregation (FLEA) iii approach, this research shows that Federated Affective Computing makes it possible for affect-enabled systems to learn quickly and perform accurate decision-making when limited data is available for training. More importantly, however, this research shows that by adopting Federated Affective Computing, affect-enabled platforms can be effectively deployed in sensitive scenarios, such as pain management. Through the adoption of Federated Affective Computing as part of the Morpheus case study, this research shows the adoption of Federated Affective Computing to be an effective tool that can take the administration of Distraction Therapy using technology-based approaches to the next level. The current state-of-the-art approach, involving non- affect-enabled software to distract patients from pain, recorded a drop of 50% in pain scores amongst a test population. This research shows that the adoption of the presented Federated Affective Computing approach led to an average 17% improvement in patient pain tolerance over the current state of the art and, in some cases, even reaching up to a 30% increase. Description: Ph.D.(Melit.) 2022-01-01T00:00:00Z /library/oar/handle/123456789/137285 Title: Highlight detection in live streams using audience reactions with transformer language models Abstract: Livestreaming of e-sports events has become very popular in recent years, with millions of people watching livestreams of competitions and commenting synchronously in chat rooms. As an effect of the rise of e-sports, there is a demand for match highlight videos. These videos, which consist of the most exciting moments in a match, help followers of the sport to stay up-to-date with or relive past games. Since their manual creation is time intensive, automatic and semi-automatic approaches for highlight detection in live streams have been devised. In this work, we suggest a novel transformer based approach to highlight detection. We employ the audience reactions found in live stream chat in order to find gripping segments of livestreams. To this end, we suggest an approach which combines contextual transformer embeddings with additional temporal features of the chat. We pre-train a language model for the domain of live stream chat in the game League of Legends and employ it on this task. For training this transformer language model, we collect a corpus from a popular livestreaming platform which contains audience reactions to competitive League of Legends matches. With our new model, we achieve an improvement over the state of the art of 0.01 f-score. We provide a new corpus for the domain and make available our pre-trained language model, which we call TwitchLeagueBert. Description: M.Sc. (HLST)(Melit.) 2022-01-01T00:00:00Z /library/oar/handle/123456789/137284 Title: On the cusp of comprehensibility : can language models distinguish between metaphors and nonsense? Abstract: Utterly creative texts can sometimes be difficult to understand, balancing on the edge of comprehensibility. However, good language skills and common sense allow advanced language users both to interpret creative texts and to reject some linguistic input as nonsense. The goal of this thesis is to evaluate whether the current language models are also able to make the distinction between creative language use, namely (unconventional) metaphors, and nonsense. To test this, mean rank and pseudo-log-likelihood score (PLL) of metaphorical and nonsensical sentences were computed, and several pre-trained models (BERT, RoBERTa) were fine-tuned for binary classification between the two categories. There was a significant difference between the categories in the mean ranks and PPL scores, and the classifier reached around 70.0% - 85.5% accuracy, which is close to the 87% accuracy of the human baseline. The satisfactory performance seems to signal that it is already possible to train the current language models to distinguish between metaphors and nonsense. This raises further questions on the characteristics of metaphorical and nonsensical sentences which allow the successful classification. Description: M.Sc. (HLST)(Melit.) 2022-01-01T00:00:00Z