Design Patterns in a Machine Learning Framework for Medical Diagnostics

摘要

This work presents the use of creational and structural design patterns in a generalized software framework intended to alleviate some of the difficulties in employing machine learning models in clinical applications. The design entails a configurable pipeline that supports not only the experimentation and development of diagnostic machine learning algorithms, but also their transition into production level systems. The resulting framework provides the foundation for the development of unique tools by both novice and expert users.