Modelling stochastic time delay for regression analysis

2021/11/11 by Juan Camilo Orduz, Orduz, Juan Camilo, Aaron Pickering +1
Engineering · Computer Science · #Control Systems and Identification #Neural Networks and Applications #Fault Detection and Control Systems

paper · pdf · doi:10.48550/arxiv.2111.06403

Abstract

Systems with stochastic time delay between the input and output present a number of unique challenges. Time domain noise leads to irregular alignments, obfuscates relationships and attenuates inferred coefficients. To handle these challenges, we introduce a maximum likelihood regression model that regards stochastic time delay as an "error" in the time domain. For a certain subset of problems, by modelling both prediction and time errors it is possible to outperform traditional models. Through a simulated experiment of a univariate problem, we demonstrate results that significantly improve upon Ordinary Least Squares (OLS) regression.

Citations

• The Proof and Measurement of Association between Two Things
• Investigating Causal Relations by Econometric Models and Cross-spectral Methods
• Deep Transformer Models for Time Series Forecasting: The Influenza Prevalence Case

Related

• Dynamic linear regression models for forecasting time series with semi long memory errors
• Optimal time delay embedding for nonlinear time series modeling
• On Multi-step Estimation of Delay for SDE
• Time delay estimation using phase data
• Inference of time delay in stochastic systems