Falsifiable implies Learnable

TL;DR

This paper establishes a fundamental link between falsifiability and learnability, proving that any falsifiable theory can be learned effectively in various prediction settings.

Contribution

It provides the first formal proof that falsifiability guarantees learnability across multiple learning paradigms.

Findings

Falsifiability implies learnability in statistical learning and sequential prediction.

Theorem extends to universal induction.

Falsifiable theories admit optimal prediction strategies.

Abstract

The paper demonstrates that falsifiability is fundamental to learning. We prove the following theorem for statistical learning and sequential prediction: If a theory is falsifiable then it is learnable -- i.e. admits a strategy that predicts optimally. An analogous result is shown for universal induction.