Hassan
Mousaid,
PhD
candidate
David
R.
Cheriton
School
of
Computer
Science
Data scientists in healthcare need to access a data science platform that allows them to develop, train, test and deploy their machine learning algorithms. A data science platform needs to meet a set of optimistic requirements such as scalability, modularity, and interoperability.
In this presentation, we talk about the different architectural components that make up a scalable and modular data science platform. Since there is no artificial intelligence (AI) without an information architecture (IA), we also address different data standards, data modeling naming conventions and API guidelines that the platform needs to comply with to achieve the interoperability optimistic requirement.