Capturing Dependencies within Machine Learning via a Formal Process Model
This work addresses the need for structured and formal process models in ML development, which is incremental as it builds on existing literature to provide a consistent framework.
The authors tackled the challenge of formalizing the development process of machine learning models by defining a comprehensive software development process model that includes tasks, artifacts, and specifications, and enables ML to be formulated as a meta-optimization problem for potential use in self-adaptive systems.
The development of Machine Learning (ML) models is more than just a special case of software development (SD): ML models acquire properties and fulfill requirements even without direct human interaction in a seemingly uncontrollable manner. Nonetheless, the underlying processes can be described in a formal way. We define a comprehensive SD process model for ML that encompasses most tasks and artifacts described in the literature in a consistent way. In addition to the production of the necessary artifacts, we also focus on generating and validating fitting descriptions in the form of specifications. We stress the importance of further evolving the ML model throughout its life-cycle even after initial training and testing. Thus, we provide various interaction points with standard SD processes in which ML often is an encapsulated task. Further, our SD process model allows to formulate ML as a (meta-) optimization problem. If automated rigorously, it can be used to realize self-adaptive autonomous systems. Finally, our SD process model features a description of time that allows to reason about the progress within ML development processes. This might lead to further applications of formal methods within the field of ML.