type project date 2024-02-23 context AI for Bioinformatics project, UNIMORE
Ovarian Cancer Survival Prediction
A project aimed at improving cancer-patient outcome prediction by integrating multiple data sources with Graph Neural Networks.
What I did:
- Data integration — collected and preprocessed heterogeneous sources, including genomic data and clinical records, into a single dataset.
- Graph Neural Networks — implemented GNNs with attention layers to model the relationships between data points; graphs are a natural fit for capturing the dependencies and interactions within this kind of data.