Home // SIGNAL 2025, The Tenth International Conference on Advances in Signal, Image and Video Processing // View article
Authors:
Pavel Loskot
Keywords: Linear model; Mean-square error; Minecraft; Quan- tization; Parameter estimation; System identification.
Abstract:
The best selling computer game of all times, Minecraft, represents the world as discrete blocks. The Minecraft-like worlds may be unknowingly created by many mathematical models of the real-world systems, when their inputs and outputs are discretized. This paper investigates system modeling and identification with noisy, discretized, but otherwise static inputs and outputs. Such a scenario occurs, for example, when configuring and measuring the system is time-consuming and costly. The task is to infer the model parameters from a limited number of input-output measurements. It is shown that, in this setting, the traditional least-squares model fitting is ineffective. A better strategy is to first accurately estimate the static input and output values, and then obtain the model parameters by inverting the model numerically by solving an underlying set of equations for the same number of unknown model parameters. These results have direct implications on creating and interpreting mathematical models of systems, and even physical laws, when the noisy measurements are implicitly or explicitly discretized.
Pages: 40 to 45
Copyright: Copyright (c) IARIA, 2025
Publication date: March 9, 2025
Published in: conference
ISSN: 2519-8432
ISBN: 978-1-68558-245-6
Location: Lisbon, Portugal
Dates: from March 9, 2025 to March 13, 2025