Sufficiently Accurate Model Learning for Planning

TL;DR

This paper introduces a constrained model learning approach that incorporates prior knowledge via constraints, improving the accuracy and task-specific suitability of models for planning in dynamical systems.

Contribution

It presents the Sufficiently Accurate model learning method with theoretical guarantees on solution quality based on problem parameters.

Findings

The approach improves model accuracy for planning tasks.

Theoretical bounds relate solution quality to model parameters.

Constraints help focus model capacity on important system aspects.

Abstract

Data driven models of dynamical systems help planners and controllers to provide more precise and accurate motions. Most model learning algorithms will try to minimize a loss function between the observed data and the model's predictions. This can be improved using prior knowledge about the task at hand, which can be encoded in the form of constraints. This turns the unconstrained model learning problem into a constrained one. These constraints allow models with finite capacity to focus their expressive power on important aspects of the system. This can lead to models that are better suited for certain tasks. This paper introduces the constrained Sufficiently Accurate model learning approach, provides examples of such problems, and presents a theorem on how close some approximate solutions can be. The approximate solution quality will depend on the function parameterization, loss and constraint function smoothness, and the number of samples in model learning.