Learn how to validate estimated models through various techniques in this 15-minute video tutorial from the System Identification with Julia series. Explore different validation methods including train/test split validation, frequency-domain estimation, impulse response comparisons, and residual analysis. Master the process of describing and validating data, understanding estimated impulse responses, and performing model fitting using JuliaSim and ControlSystemIdentification tools. Follow along with provided notebook files and comprehensive documentation while discovering practical applications in dynamical model estimation using input-output data and the Julia programming language. Access additional resources on validation techniques, coherence functions, and enterprise solutions for modeling, simulation and control through detailed documentation links and platform-specific guidance.
System Identification with Julia: Model Validation - Part 7
Overview
Syllabus
Validation
Data description
Estimated impulse response
Model fitting and train/test split
Validation
Frequency-domain estimate
Compare impulse responses
Residual analysis
Summary
Taught by
JuliaHub
Related Courses
-
Subspace-Based System Identification in Julia - Part 8
-
System Identification with Julia: Experiments and Excitation - Part 6
-
System Identification with Julia: Prediction-Error Method for Linear Models - Part 4
-
Prediction vs. Simulation in System Identification with Julia - Part 3
-
System Identification with Julia: Introduction to Dynamical Models - Part 1
-
System Identification with Julia: Prefiltering Data for Model Estimation - Part 5
