DCU Expo 2024 - Final Year Projects

47 51. EnhancingDisease Classification in Raman Spectra of Exosomes Through Feature Engineering and Graph Databases This project explores howmachine classification and Raman spectroscopy can be combined to detect disease in a patient based on exosomes, small particles emitted by cells. Raman spectroscopy yields a unique spectral ‘fingerprint’ of a substance which outlines its molecular structure. Amachine learning algorithm is trained to differentiate between the spectra of healthy and diseased exosomes. The performance and explainability of this model are then enhanced using feature engineering. Peaks in the Raman spectra represent specific molecular bonds, allowing for the creation of a refined, more explainable feature set. Additionally, a graph databases construction can capture the non-uniform nature of the data, providing a source of additional features based on graph metrics. Student Programme Data Science Project Area Artificial Intelligence, Data Analytics, Digital Signal Processing, Spectral Analysis, Graph Databases Project Technology Digital Signal Processing, Python, Machine Learning, Neo4j Student Name(s) Edward Bolger  |  Stan Goodwin Email edward.bolger25@mail.dcu.ie   |   stan.goodwin6@mail.dcu.ie Supervisor Prof Mark Roantree 52. DegreeDive DegreeDive aims to transform the college selection process with an aptitude testing system resulting in personalised learning recommendations. It provides the student with the best-suited level 8 degrees to them in their chosen Irish colleges and within their predicted points range. It will aid students in aligning their academic and personal interests with suitable college courses. This tool helps students navigate the complexity of selecting a course, aiming to help students fill out their CAO. The application also features a CAO points calculator and a Grade Goal Setting function, enabling students to set achievable targets. DegreeDive aims to reduce dropout rates and empower students with informed choices, facilitating a smoother educational journey. Student Programme Enterprise Computing Project Area Automation, DataAnalytics, Educational, Information Retrieval, WebApplication Project Technology CSS, Docker, HTML5, JavaScript, Python, Django Student Name(s) Ellen Kenny  |  Grace Tyrrell Email ellen.kenny35@mail.dcu.ie   |   grace.tyrrell5@mail.dcu.ie Supervisor Prof Cathal Gurrin 53. Oscillatory FlowBehaviour inWater-Air Hourglasses This project investigates the intriguing oscillatory flow observed during the emptying of bottles, shedding light on the complex fluid dynamics underlying this seemingly simple process. Through experimentation and analysis, the role of gauge pressure, bottle diameter, bottle-neck diameter, initial water height and how they affect flow patterns is investigated. High-speed imaging and pressure sensors facilitate the graphing and comparison of key variables and flow parameters. In understanding the effects of these, the process of bottle-emptying can be optimised to reduce emptying time and increase flow velocity, which can prove important in an industry where optimising flow behaviours can enhance product quality and efficiency. Student Programme Mechanical andManufacturing Engineering (Year 5) Project Area FluidMechanics, Mechanical Design andManufacture Project Technology Matlab, Solidworks Student Name(s) AndrewGavin Email andrew.gavin22@mail.dcu.ie Supervisor Dr CornéMuilwijk