Learnable: Theory vs Applications
This addresses the gap between theoretical machine learning and practical applications, highlighting an incremental critique of existing frameworks.
The paper compares Applied learning and Agnostic PAC learning, showing that PAC theory can solve Applied learning under certain conditions but requires impractically large training sets, suggesting a need to align theory with practitioner experience.
Two different views on machine learning problem: Applied learning (machine learning with business applications) and Agnostic PAC learning are formalized and compared here. I show that, under some conditions, the theory of PAC Learnable provides a way to solve the Applied learning problem. However, the theory requires to have the training sets so large, that it would make the learning practically useless. I suggest shedding some theoretical misconceptions about learning to make the theory more aligned with the needs and experience of practitioners.