On the design of autonomous agents from multiple data sources

TL;DR

This paper presents a data-driven optimal control framework for autonomous agents to dynamically select information from multiple sources, optimizing task performance and providing regret bounds with simulation validation.

Contribution

It introduces a novel formulation of the agent data selection problem as an optimal control problem with explicit solutions and regret analysis.

Findings

Explicit solution for data source selection in optimal control

Regret upper bounds for the agent's decision process

Simulation results validating the theoretical approach

Abstract

This paper is concerned with the problem of designing agents able to dynamically select information from multiple data sources in order to tackle tasks that involve tracking a target behavior while optimizing a reward. We formulate this problem as a data-driven optimal control problem with integer decision variables and give an explicit expression for its solution. The solution determines how (and when) the data from the sources should be used by the agent. We also formalize a notion of agent's regret and, by relaxing the problem, give a regret upper bound. Simulations complement the results.