Inference of Some Contributions To Nonlinear Valuation In Regression Models

Abstract

To explain the connection between a response and an independent variable, regression models are often employed in various applied sciences fields. The maximum likelihood estimators are often used to make statistical conclusions about the regression parameters (MLE). It is typical practise to linearize a nonlinear function around the real parameter and then utilise large sample theory to generate MLE standard errors. Here we show that the asymptotic Wald confidence intervals generated by computer simulation do not meet the required confidence levels. They may overestimate the real nominal level and become too generous as a result. As a result, caution should be used when utilizing the conventional linearized MLE error and confidence intervals that go along with it. When calculating MLE standard errors, it is typical to linearize a nonlinear function around the parameter in question and then utilise large sample theory to do so. We demonstrate in this study that the asymptotic Wald confidence intervals that come from computer simulations fall short of the needed levels of confidence. In other words, they may overestimate the true nominal level and overextend themselves as a consequence. As a result, caution should be used when utilizing the conventional linearized MLE error and confidence intervals that go along with it. The lightweight, high-strain, and muscle-like characteristics of new developing classes of next-generation soft-smart actuators will have a significant effect on robotics, orthotics, and prosthetics. Due to their variable compliance, these actuators, like muscle, may serve several purposes, such as acting as both an actuator and a brake. The dielectric elastomer actuator is an essential kind of soft actuator (DEA). However, owing of their non-linear and time-varying dynamics, DEAs are very challenging to manage. The creation of methods for systems level modeling and analysis, on which this thesis is focused, is a critical step in the progress of this technology.