Amortised Experimental Design and Parameter Estimation for User Models of Pointing

TL;DR

This paper introduces an amortised experimental design method that trains a policy using simulated user interactions to efficiently estimate user model parameters, reducing the need for extensive real user data.

Contribution

It presents a novel approach that amortises the computational cost of experiment design by leveraging simulated agents to optimize data collection for user models.

Findings

Efficient parameter estimation for pointing models.

Reduced reliance on large-scale human data.

Effective for progressively complex user models.

Abstract

User models play an important role in interaction design, supporting automation of interaction design choices. In order to do so, model parameters must be estimated from user data. While very large amounts of user data are sometimes required, recent research has shown how experiments can be designed so as to gather data and infer parameters as efficiently as possible, thereby minimising the data requirement. In the current article, we investigate a variant of these methods that amortises the computational cost of designing experiments by training a policy for choosing experimental designs with simulated participants. Our solution learns which experiments provide the most useful data for parameter estimation by interacting with in-silico agents sampled from the model space thereby using synthetic data rather than vast amounts of human data. The approach is demonstrated for three progressively complex models of pointing.