Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
October 10, 2025Open Access

An Approximate Bayesian Approach to Optimal Input Signal Design for System Identification

View Full Paper
Ask AI
Bookmark
Share

Authors

PBPiotr BaniaJagiellonian UniversityAWAnna WójcikJagiellonian University

Discussion

Loading...

Member takes

Overview

This paper demonstrates how a Bayesian method for input signal design optimizes mutual information in system identification, addressing model uncertainties.

Key Points

  • The proposed Bayesian method optimizes input signals to enhance mutual information in system identification, reducing estimation errors.
  • Maximizing a tractable mutual information lower bound allows for input signal design that mitigates computational challenges in high-dimensional spaces.
  • The approach effectively handles quasi-linear stochastic dynamical systems by efficiently inverting large covariance matrices.
  • Comparisons show that the Bayesian method outperforms classical and semi-Bayesian approaches in accuracy and feasibility for extensive experiments.

Cite This Study

Bania et al. (2025) studied this question.

synapsesocial.com/papers/68e861907ef2f04ca37e40d7https://doi.org/10.20944/preprints202509.1390.v3
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1A generalized Bayesian framework for maximizing information gain and model selection2026
  2. 2Optimal Input Design for Fractional-Order System Identification Using an LMI-Based Frequency Error Criterion2025 · 1 citations
  3. 3Dynamic importance learning using fisher information gain for nonlinear system identification2024
  4. 4An information-matching approach to optimal experimental design and active learning2026
  5. 5Coherence-based Input Design for Sparse System Identification2024