Simple and Fast Algorithms for Interactive Machine Learning with Random Counter-examples

TL;DR

This paper introduces simple, efficient algorithms for interactive learning from random counter-examples, achieving near-optimal learning times and outperforming previous methods in practical scenarios.

Contribution

It presents deterministic and randomized algorithms for non-binary concept learning in the LRC model, solving an open problem and generalizing prior results.

Findings

Both algorithms have an $ ext{O}(\log|H|)$ average learning time.

Expected learning time is bounded by $ ext{O}(rac{1}{\epsilon}\log{rac{|H|}{\delta}})$.

Algorithms outperform theoretical bounds in practical simulations.

Abstract

This work describes simple and efficient algorithms for interactively learning non-binary concepts in the learning from random counter-examples (LRC) model. Here, learning takes place from random counter-examples that the learner receives in response to their proper equivalence queries. In this context, the learning time is defined as the number of counter-examples needed by the learner to identify the target concept. Such learning is particularly suited for online ranking, classification, clustering, etc., where machine learning models must be used before they are fully trained. We provide two simple LRC algorithms, deterministic and randomized, for exactly learning non-binary target concepts for any concept class $H$. We show that both of these algorithms have an $\mathcal{O}(\log{}|H|)$ asymptotically optimal average learning time. This solves an open problem on the existence of an efficient LRC randomized algorithm while simplifying and generalizing previous results. We also show that the expected learning time of any arbitrary LRC algorithm can be upper bounded by $\mathcal{O}(\frac{1}{\epsilon}\log{\frac{|H|}{\delta}})$, where $\epsilon$ and $\delta$ are the allowed learning error and failure probability respectively. This shows that LRC interactive learning is at least as efficient as non-interactive Probably Approximately Correct (PAC) learning. Our simulations show that in practice, these algorithms outperform their theoretical bounds.