MEskink: Generalized Maximum Entropy for estimating the smooth...

In woraphonyamaka/GMEreg: Generalized Maximum Entropy regression

Description

GME inference method for the smooth transition kink regression model with under kink point. The advantage of GME method is that it is robust even when we have ill-posed or ill-conditioned problems, and thus, it has higher estimation accuracy and robustness, especially when the probability distribution of errors is unknown.

Usage

MEskink(y,x,number,Z,V)

Arguments

y	dependent variable
x	One dimension of dependent variable
number	number of supports i.e. "3", "5" and "7
Z	bound of coefficient
V	bound of error

Details

Entropy refers to the amount of uncertainty represented by a discrete probability distribution. The maximum entropy method was proposed by Jaynes (1957) and developed in the early 1990s by Golan, Judge, and Miller (1996) for estimating the unknown probabilities of a discrete probability distribution. This estimator uses the entropy-information measure of Shannon (1948) to recover those unknown probability distributions of underdetermined problems. This function is a simple estimation function for one covariate.

Value

beta	intercept,beta_regime1,beta_regime2
threshold	kink point
smooth	kink point
Maxent	Maximum entropy

Author(s)

Dr.Woraphon Yamaka

References

Golan, A., Judge, G. G., & Miller, D. (1996). Maximum entropy econometrics. Iowa State University, Department of Economics.

Jaynes, E. T. (1957). Information theory and statistical mechanics. Physical review, 106(4), 620.

Maneejuk, P. and Yamaka, W. (2020). Entropy Inference in Smooth Transition Kink Regression

Examples

library("Rsolnp")
set.seed(1)
n=100
thres=3
gam=1.2
e=rnorm(n)
x=rnorm(n,thres,5)
alpha=c(0.5,1,-1)

y=alpha[1]+(alpha[2]*(x*(1-logis(gam,x,thres))))+(alpha[3]*(x*(logis(gam,x,thres))))+e

MEskink(y,x,number="5",Z=10,V=5)

woraphonyamaka/GMEreg documentation built on July 28, 2020, 9:59 a.m.