A Pattern Language for Machine Learning Tasks
This provides a foundational framework for unifying and optimizing machine learning tasks across domains, potentially impacting all of ML/AI.
The paper tackles the problem of formalizing machine learning tasks as equality constraints on learner composites, proposing a graphical mathematics to unify approaches, design behaviors model-agnostically, and import theoretical insights. As a proof-of-concept, they implement a manipulator task that edits input data to have desired attributes, achieving this end-to-end without custom architectures or adversarial training.
We formalise the essential data of objective functions as equality constraints on composites of learners. We call these constraints "tasks", and we investigate the idealised view that such tasks determine model behaviours. We develop a flowchart-like graphical mathematics for tasks that allows us to; (1) offer a unified perspective of approaches in machine learning across domains; (2) design and optimise desired behaviours model-agnostically; and (3) import insights from theoretical computer science into practical machine learning. As a proof-of-concept of the potential practical impact of our theoretical framework, we exhibit and implement a novel "manipulator" task that minimally edits input data to have a desired attribute. Our model-agnostic approach achieves this end-to-end, and without the need for custom architectures, adversarial training, random sampling, or interventions on the data, hence enabling capable, small-scale, and training-stable models.