My overarching interests focus on the application and uptake of machine and deep learning models in clinical settings. In particular I am interested in driving AI towards mental health disorders, and developing sensible, usable, and practical tools to support clinical decision making. A recent focus of mine has been on the field of explainability and transparency for ML models and integrating modern methods into end-to-end pipeline with the goal of avoiding opaque, black-box style models. Current projects pertain to Natural Language Processing (NLP) and model explain-ability with the application to clinical free text data contained within NHS electronic health records.
To further encourage transparency in my work, I try to ensure my code is designed with scalability and reproducibility as priorities. Where possible, code and data used for projects will be packaged into a high quality github repository and made publically available.