Online Learning with Improving Agents: Multiclass, Budgeted Agents and Bandit Learners

TL;DR

This paper explores a model of online learning where agents can modify features to achieve better labels, extending theoretical understanding to multiclass, bandit feedback, and cost modeling scenarios.

Contribution

It provides new combinatorial dimensions for characterizing online learnability, extending prior work to multiclass, bandit feedback, and cost-aware settings.

Findings

Derived combinatorial dimensions for learnability

Extended analysis to multiclass scenarios

Analyzed bandit feedback and cost modeling

Abstract

We investigate the recently introduced model of learning with improvements, where agents are allowed to make small changes to their feature values to be warranted a more desirable label. We extensively extend previously published results by providing combinatorial dimensions that characterize online learnability in this model, by analyzing the multiclass setup, learnability in a bandit feedback setup, modeling agents' cost for making improvements and more.