Page 269 - Contributed Paper Session (CPS) - Volume 7
P. 269
CPS2094 Yoshimitsu Morinishi et al.
5. Convert explanatory variable to categorical variable and Convert to
polychoric Correlation Matrix: Considering that the objective variable
used this time is 01 data, in order to increase the goodness of fitting, all
the variables to be used are categorical variables After that, we
calculated the polychoric correlation matrix which is the correlation
matrix of the categorical variables.
6. Formulation of a causal model for each bat by SEM (structural equation
model): Using the SEM (structural equation model) function of JUSE ·
Stats Works package of Nikka Giken Statistics usiness analysis package,
batting strategy We verify the causal effect of the fly ball out of each
swing speed verified in this research by verifying the causal effect
against the out of the fly ball and comparing
4. Outcome
From the result of the causal model, the significance of the contribution of
each explanatory variable (both latent variable and observed variable) to the
objective variable was confirmed together with the concept. It is thought that
it is effective in explanation on the site that it can explain by not only a
prediction model as a simple black box but also a model unique to SEM.
Moreover, as the result of SEM model was not sufficiently adapted, we
adopted logistic regression analysis as the final model, and we were able to
verify the effectiveness of "flyball revolution" at NPB. Specifically, I found the
following.
256 | I S I W S C 2 0 1 9