A Scheme-Driven Approach to Learning Programs from Input/Output Equations

TL;DR

This paper presents a fast, scheme-driven method for learning recursive function definitions from input/output equations, demonstrated through a Prolog prototype, with potential for further refinement.

Contribution

It introduces a novel, efficient approach to learn recursive programs from input/output data within a term-rewriting framework, focusing on structurally recursive definitions.

Findings

The approach effectively learns recursive functions from input/output equations.

The method produces definitions close to intuitive expectations.

A Prolog prototype demonstrates practical implementation.

Abstract

We describe an approach to learn, in a term-rewriting setting, function definitions from input/output equations. By confining ourselves to structurally recursive definitions we obtain a fairly fast learning algorithm that often yields definitions close to intuitive expectations. We provide a Prolog prototype implementation of our approach, and indicate open issues of further investigation.