Combining deep learning and ontologies
The project aims to identify applications of deep learning to classification problems involving ontologies, and apply these methods to biological and biomedical datasets. The main challenges of the project are (1) to develop methods that can effectively be applied both to unstructured and structured data and classify instances into classes from ontologies, and (2) to utilize data already structured with ontologies effectively in classification and regression problems. The students will be provided with real-world several datasets and are expected to implement and evaluate deep learning approaches on these datasets. To evaluate the methods on a large scale, students will have access to one of the compute clusters at KAUST.