Your data analysis solution
Name is required.
Email address is required.
Invalid email address
Answer is required.
Exceeding max length of 5KB

PCA

By Michael Ngai | Apr 16, 2018 09:42PM CEST

Hi everyone,

If we want to extract a subsample of insurance data to do the PCA, what kind of sampling method do you recommend (e.g. Random without/with replacement, Systematic with random start, Systematic Centered, Random Stratified)?

Based on our understanding, Gini Coefficient is not supported in XLSTAT. If we want to evaluate PCA models under different rotations or compare PCA model with GLM model in XLSTAT, what should we do instead?

If we want to calculate Gini Coefficients for the PCA models in excel based on the first principle, how can we do that?

Thanks and regards,

Michael

Up

0

Down

By Sébastien | Apr 18, 2018 03:39PM CEST | XLSTAT Agent

Hello,

Choosing a sampling method depends on the assumptions about the initial population.
As an example systematic sampling is to be applied only if the given population is logically homogeneous, because systematic sample units are uniformly distributed over the population.
Another classical method is to selecting subjects completely at random (random sampling with/without replacement) from the larger population; this yields to a sample that is representative of the group being studied. This method also insures lowest sampling error if the only assumption is that your sample is a good representation of the population.

Concerning the Gini Coefficients, unfortunately, there is no functionality in XLSTAT that allows you to compute that.

You canbalso have a look at the following guide linked below to know which statistical model should you choose :
https://help.xlstat.com/customer/en/portal/articles/2062461-quel-outil-de-modelisation-statistique-choisir?b_id=9283

Regards,

Post Your Public Answer

Contact Us

1c26995d494fb3061dd0ae8571ffc0a4@xlstat.desk-mail.com
https://cdn.desk.com/
false
desk
Loading
seconds ago
a minute ago
minutes ago
an hour ago
hours ago
a day ago
days ago
about
false
Invalid characters found
/customer/portal/articles/autocomplete
9283