DCU Expo 2024 - Final Year Projects
31 3. Lower Limb Sports Injury Prediction and Prevention Our project uses machine learning to predict and prevent lower limb injury in American football. We use Decision Tree, Random Forest, and XGBoost models on a play-by-play NFL dataset fromKaggle, incorporating player and injury details. We also incorporate K-fold cross-validation to ensure the model’s robustness. To showcase our results and findings, we have developed a user-friendly web application using React.js. This web app will enable stakeholders, such as trainers andmedical professionals, to easily access and understand the model’s predictions and insights, and therefore taking a proactive approach to injury prevention, safeguarding athletes’ well-being and optimising their performance. Student Programme Computer Applications Project Area Artificial Intelligence, Data Analytics, DataMining Project Technology JavaScript, Python, React.js, Machine Learning Student Name(s) Dylan Andrew Bagnall | Lorcan Dunne Email dylan.bagnall2@mail.dcu.ie | lorcan.dunne47@mail.dcu.ie Supervisor Prof Mark Roantree 4. Implementation and Performance Analysis of the RV32E Base RISC-VArchitecture This project investigates the functionality of the RISC-V 32E architecture through the implementation of a single-cycle, multi-cycle and pipelined processor. RISC-V is an open-source alternative to existing industry ISAs such as x86 and holds promising use cases in FPGA engineering and other embedded applications. Through the use of SystemVerilog hardware development language, the three processors are created as working FPGA design schematics and their advantages and disadvantages analysed. Student Programme Mechatronic Engineering (Year 5) Project Area Circuit Modeling, Device Design, Embedded Systems, Simulation Project Technology FPGA Student Name(s) Pietre O’Sullivan Email pietre.osullivan42@mail.dcu.ie Supervisor Dr XiaojunWang 5. Price Profile Computer Application for EVCharging Stations As demand for electric vehicles (EVs) continues to grow, management of EV charging stations needs to become more data-driven. The motivation for our project is tomaximise revenue for station owners during peak times and help themattract customers during off-peak times. To achieve this goal, an application capable of generating hourly pricing profiles for charging stations using innovative demand- proportional pricingmethods was developed. In the application, demand is modelled usingmachine learning/statistical algorithms, such as LSTM, GRU, SARIMA and TBATS. The projected demand is then used to generate the pricing profile. Various statistics and plots are generated on the historical and forecasted data and displayed to the user on a dashboard. Student Programme Data Science Project Area Artificial Intelligence, Data Analytics, Simulation, Statistical Analysis, Forecasting Project Technology Python, Machine Learning, Hive, Pig Student Name(s) Krzysztof Baran | Xi Zhang Email krzysztof.baran2@mail.dcu.ie | xi.zhang28@mail.dcu.ie Supervisor Dr Mohammed Amine Togou
Made with FlippingBook
RkJQdWJsaXNoZXIy MTQzNDk=