Bezier Correction Coefficient _ machine learning--notes

Meaning of different variance

S=σ (X−x¯) 2n−−−−−−√ meaning: Represents the variance of a population dataset.
S=σ (X−x¯) 2n−1−−−−−−√ meaning: Represents the variance of a sample. Why would you do that?

The reason is: for example, in the Gaussian distribution, we take a part of the sample, using the variance of the sample to represent the variance of the large sample dataset that satisfies the Gaussian distribution. Since the sample is mainly in the vicinity of the X=u center value, then if the sample is S=σ (X−x¯) 2n−−−−−−√ the variance, then the predicted variance must be less than the variance of the large dataset (because the Gaussian distribution edge extracts very little data). In order to compensate for this shortcoming, we change the N of the formula to n-1 to increase the value of the variance. This method is called the Bezier correction factor.
See the video specifically