The network structural equation modeling conducts a network
statistical analysis on a data frame of coincident observations of
multiple continuous variables [1].
It builds a pathway model by exploring a pool of domain knowledge guided
candidate statistical relationships between each of the variable pairs,
selecting the 'best fit' on the basis of a specific criteria such as
adjusted r-squared value.
This material is based upon work supported by the U.S. National Science
Foundation Award EEC-2052776 and EEC-2052662 for the MDS-Rely IUCRC Center,
under the NSF Solicitation:
NSF 20-570 Industry-University Cooperative Research Centers Program
[1] Bruckman, Laura S., Nicholas R. Wheeler, Junheng Ma, Ethan Wang,
Carl K. Wang, Ivan Chou, Jiayang Sun, and Roger H. French. (2013)
The R package 'netSEM' conducts a net-SEM statistical analysis (network structural equation modeling) on a data frame of coincident observations of multiple continuous variables. Principle 1 generates an inferential model through pairwise correlation of variables based on the Markovian Spirit. Principle 2 provides a predictive model through multiple regression with the model complexity and performance evaluated using either Akaike Information Criterion (AIC) or Bayesian Information Criterion (BIC) specified by the user.
This is a simple example for generating degradation models using netSEM principle 1 and principle 2.
# Load in netSEM library
library(netSEM)
# Load in example data
data(acrylic)
# Perform principle 1 and principle 2
acrylic_p1 <- netSEMp1(acrylic)
acrylic_p2 <- netSEMp2(acrylic, criterion = "AIC") #AIC by default
# Plotting netSEM diagrams
plot(acrylic_p1, cutoff = c(0.3,0.6,0.8))
plot(acrylic_p2, cutoff = c(0.3,0.6,0.8))