MSc Operational Research and Applied Statistics MSc Operational Research, Applied Statistics and Risk MSc Data Science and Analytics MAT099 (MSc Dissertation) Project Brief Project title: Machine Learning & Data Analysis Number of students to be allocated to project: Up to 6 Sponsor organisation: Cardiff University Sponsor Supervisor: Bertrand Gauthier University Contact: Joanna Emery (EmeryJL4@cardiff.ac.uk) The Organisation: NA Background: Machine Learning (ML) and Data Analysis (DA). Objectives: This project aims at exploring the application of various DA and ML Learning (ML) techniques for the analysis and data-based modelisation of challenging real-life problems. A dataset and its associated problematic will be provided to the students. The considered problem will be related to prediction of the properties of some nano materials; the data are supplied by the School of Chemistry of Cardiff University. (Depending on the will of the students, some other dataset and problematics might potentially be considered). Approach: A general statistical analysis will first be performed to gain an overall understanding of the properties of the dataset; the students will in parallel explore the relevant literature and identify a set of ad hoc ML techniques for the modelisation of some specific features of the dataset. The selected techniques will next be implemented, and their efficiency assessed. If time allows, some optimisation techniques will be applied to the obtained models in order to help identifying some nano-material configurations of interest. Deliverables: Report and sample of the codes produced during the project. Location of the project: NA Key computing skills: Computations will preferably be performed in R or Python (using Jupyter notebooks for instance), but any other adequate programming language and environment can also be considered. Page 1 of 2 MSc Operational Research and Applied Statistics MSc Operational Research, Applied Statistics and Risk MSc Data Science and Analytics Other key student competencies: Curiosity, seriousness and resourcefulness. Data availability: Yes. Any other comments: No. Page 2 of 2